{"id":58,"date":"2020-09-03T16:38:55","date_gmt":"2020-09-03T13:38:55","guid":{"rendered":"http:\/\/acikkitap.com.tr\/oaek\/chapter\/bolum-10-dogal-dil-isleme-ve-ogrenme-analitigi\/"},"modified":"2020-09-03T16:38:55","modified_gmt":"2020-09-03T13:38:55","slug":"bolum-10-dogal-dil-isleme-ve-ogrenme-analitigi","status":"publish","type":"chapter","link":"https:\/\/acikkitap.com.tr\/oaek\/chapter\/bolum-10-dogal-dil-isleme-ve-ogrenme-analitigi\/","title":{"raw":"B\u00f6l\u00fcm 10 Do\u011fal Dil \u0130\u015fleme ve \u00d6\u011frenme Analiti\u011fi","rendered":"B\u00f6l\u00fcm 10 Do\u011fal Dil \u0130\u015fleme ve \u00d6\u011frenme Analiti\u011fi"},"content":{"raw":"\n

<\/a> Sidney K. D Mello<\/span><\/span><\/p>\nPsikoloji ve Bilgisayar Bilimleri ve M\u00fchendisli\u011fi B\u00f6l\u00fcmleri, Notre Dame \u00dcniversitesi, ABD<\/span><\/span>\n\nDOI: 10.18608\/hla17.010<\/span><\/span>\n

\u00d6Z<\/h2>\nBu b\u00f6l\u00fcm, duygular\u0131n \u00f6\u011frenmenin yayg\u0131nl\u0131\u011f\u0131n\u0131 ve \u00f6nemini tart\u0131\u015fmaktad\u0131r. Ke\u015fif odakl\u0131, veri g\u00fcd\u00fcml\u00fc, \u00f6\u011frenme analiti\u011fi (\u00d6A) ve e\u011fitsel veri madencili\u011fi (EVM) ile duygusal ve \u00f6\u011frenme bilimlerindeki teorik geli\u015fmeler ve metodolojileri birle\u015ftirerek kayda de\u011fer ilerleme sa\u011flanabilece\u011fini savunuyor. Bu alanlar\u0131n kesi\u015fimindeki temel, ortaya \u00e7\u0131kan ve gelecekteki ara\u015ft\u0131rma temalar\u0131 tart\u0131\u015f\u0131lmaktad\u0131r.<\/span>\n\nAnahtar kelimeler<\/span>: Etkiler, duyu\u015fsal bilim, duyu\u015fsal bilgi i\u015flem, e\u011fitsel veri madencili\u011fi<\/span>\n

Bir makalem i\u00e7in bir hakemin \u201c\u00f6nerisiyle\u201d (D'Mello, 2016), son d\u00f6nemde (benim i\u00e7in) yeni bir istatistiksel y\u00f6ntem olan genelle\u015ftirilmi\u015f toplan\u0131r karma modelleri \u00f6\u011frenmeye ba\u015flad\u0131m GTKM; McKeown ve Sneddon, 2014). GTKM bir cevap de\u011fi\u015fkenini (rezid\u00fcel) art\u0131klar aras\u0131ndaki oto d\u00fczeltmelere de\u011finerek (zaman serileri verisinde), \u00f6ng\u00f6r\u00fcsel de\u011fi\u015fkenlere ait parametrik ve parametrik olmayan d\u00fczg\u00fcn fonksiyonlar\u0131n toplan\u0131r bir birle\u015fimi ile modellemeyi ama\u00e7lar. \u0130lk ba\u015fta, bu makale i\u00e7in biraz daha fazla \u00e7al\u0131\u015fma d\u00fc\u015f\u00fcncesi beni biraz ho\u015fnutsuz k\u0131ld\u0131. Endi\u015fem son d\u00fczeltme tarihine kadar yeni bir metodu \u00f6\u011frenmek ve uygulamak i\u00e7in yeterli zaman\u0131m\u0131n olmayaca\u011f\u0131na dair d\u00fc\u015f\u00fcncemden kaynakland\u0131. Hi\u00e7bir \u015fey yapmad\u0131m. Son tarih yakla\u015f\u0131rken kayg\u0131 hafif pani\u011fe d\u00f6n\u00fc\u015ft\u00fc. Son olarak tavsiye edilen bir makaleyi indirerek GTKM'lere bakmaya karar verdim. Makale g\u00f6z al\u0131c\u0131 grafiklere sahipti, bu da merak duygumu uyand\u0131rd\u0131 ve beni daha fazla ke\u015ffetmeye motive etti. Merak, metot hakk\u0131nda daha \u00e7ok okuduk\u00e7a h\u0131zl\u0131 bir \u015fekilde ilgiye son olarak da yakla\u015f\u0131m\u0131n g\u00fcc\u00fcn\u00fc fark etti\u011fimde de heyecana d\u00f6n\u00fc\u015ft\u00fc. Bu bir \u015feyler anlam ifade etmedi\u011finde kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131<\/i><\/span> ve hayal k\u0131r\u0131kl\u0131\u011f\u0131,<\/i><\/span> neredeyse pes edecekken umutsuzluk, ilerleme kaydetti\u011fimi d\u00fc\u015f\u00fcnd\u00fc\u011f\u00fcmde umut ve son olarak da ger\u00e7ekten bir ilerleme kaydetti\u011fimde ise mutluluk ve haz gibi baz\u0131 yo\u011fun duygulara neden olarak beni teknik detaylarda bata \u00e7\u0131ka ilerlemeye motive etti. Daha sonra baz\u0131 R s\u00f6z dizimi \u00fczerinde de\u011fi\u015fiklikler yaparak metodu kendi verim \u00fczerinde uygulamaya koyuldum. Umut<\/i><\/span>, haz ve mutlulu\u011fun<\/i><\/span> aras\u0131na daha fazla hayal k\u0131r\u0131kl\u0131\u011f\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve <\/i><\/span>umutsuzluk serpi\u015ftirildi<\/i><\/span>. Sonunda hepsini \u00e7al\u0131\u015ft\u0131rd\u0131m ve sonu\u00e7lar\u0131 yazd\u0131m. Yazma ve d\u00fczeltme d\u00f6ng\u00fclerinde baz\u0131 ba\u015fka duygular daha olu\u015ftu. Sonunda ba\u015fard\u0131m. G\u00f6n\u00fcl rahatl\u0131\u011f\u0131, hafifleme<\/i><\/span> ve biraz da \u00f6v\u00fcn\u00e7<\/i><\/span> hissediyordum. Bu \u00f6rne\u011fin de g\u00f6sterdi\u011fi gibi, \u00f6\u011frenme s\u00fcreci boyunca bir duygular dip dalgas\u0131 vard\u0131r. Bu t\u00fcm \u201cbili\u015fin\u201d \u201cduygularla\u201d ili\u015fkili oldu\u011fu \u00f6\u011frenmeye has de\u011fildir. Duygular her zaman bilin\u00e7li olarak deneyimlenmeyebilirler (Ohman ve Soares, 1994) ancak yine de vard\u0131rlar ve bili\u015fi etkilerler. Ayn\u0131 zamanda, duygular bir hava bo\u015flu\u011funda olu\u015fmazlar, \u00f6\u011frenmenin sosyal kuma\u015f\u0131yla derinden sarmalanm\u0131\u015f vaziyettedirler. En temel i\u015fi \u00f6\u011frenmek olan tipik bir \u00f6\u011frenen taraf\u0131ndan hangi duygular yelpazesinin deneyimlenece\u011fini hayal etmek \u00e7ok zor de\u011fildir. Pekrun ve Stephens (2011), bunlara \u201cakademik duygular\u201d demi\u015f ve onlar\u0131 d\u00f6rt grupta s\u0131n\u0131fland\u0131rm\u0131\u015ft\u0131r. Ba\u015far\u0131 duygular\u0131 (g\u00f6n\u00fcl rahatl\u0131\u011f\u0131, endi\u015fe ve hayal k\u0131r\u0131kl\u0131\u011f\u0131), \u00f6\u011frenme etkinlikleri (\u00f6dev, bir teste girme) ve \u00e7\u0131kt\u0131larla (ba\u015far\u0131, ba\u015far\u0131s\u0131zl\u0131k) ba\u011flant\u0131l\u0131d\u0131r. Konu ba\u015fl\u0131\u011f\u0131<\/i><\/span> duygular\u0131 \u00f6\u011frenme i\u00e7eri\u011fi (klasik edebiyat okurken hik\u00e2yenin kahraman\u0131 ile duyguda\u015fl\u0131k kurmak) ile uyumludur. Sosyal duygular \u00f6v\u00fcn\u00e7, utan\u00e7 ve k\u0131skan\u00e7l\u0131k da vard\u0131r \u00e7\u00fcnk\u00fc e\u011fitim sosyal ortamlarda yer al\u0131r. Son olarak, \u00f6zg\u00fcnl\u00fckle kar\u015f\u0131la\u015f\u0131ld\u0131\u011f\u0131nda \u015fa\u015fk\u0131nl\u0131k veya bir a\u00e7mazla kar\u015f\u0131la\u015f\u0131ld\u0131\u011f\u0131nda kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 olmas\u0131 gibi epistemik<\/i><\/span> duygular bili\u015fsel s\u00fcre\u00e7lerden do\u011farlar.<\/p>\n

Duygular sadece tesad\u00fcfi de\u011fildirler veya evrimle\u015fmemi\u015ftirler (Darwin, 1872; Tracy, 2014). Duygular bilgi ile ilgili problemleri (kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131), uyar\u0131lma ile ilgili problemleri (b\u0131kk\u0131nl\u0131k), yakla\u015fan performansla ilgili meseleleri (endi\u015fe) ve kolayl\u0131kla a\u015f\u0131lamayacak zorluklar\u0131 (hayal k\u0131r\u0131kl\u0131\u011f\u0131) vurgulayarak uyar\u0131 verme i\u015flevini<\/i><\/span> g\u00f6r\u00fcrler (Schwarz, 2012). \u0130nsanlar\u0131n bir olaya de\u011feri hedef uygunlu\u011fu ve hedef e\u015fle\u015fmesi a\u00e7\u0131s\u0131ndan de\u011fer bi\u00e7ti\u011fi bir para birimi g\u00f6revi g\u00f6ren de\u011ferlendirici i\u015flevler<\/i><\/span> icra ederler. (Izard, 2010). Duygular bili\u015fsel oda\u011f\u0131 s\u0131n\u0131rlayarak ya da geni\u015fleterek olumlu duygularla daha geni\u015f, yukar\u0131dan a\u015fa\u011f\u0131ya ve \u00fcretici i\u015flemeyi (geni\u015fletilmi\u015f odak) desteklemeye (Barth ve Funke, 2010; Schwarz, 2012) k\u0131yasla, olumsuz duygularla i\u015flemenin dar, a\u015fa\u011f\u0131dan yukar\u0131ya ve odaklanm\u0131\u015f modellerini<\/i><\/span> (s\u0131n\u0131rl\u0131 odak) olu\u015fturarak ge\u00e7i\u015f i\u015flevini icra ederler (Fredrickson ve Branigan, 2005; Isen, 2008). Ger\u00e7ekten de duygular bellek, problem \u00e7\u00f6zme, karar lama ve bili\u015fin di\u011fer alanlar\u0131 \u00fczerindeki etkilerinde a\u00e7\u0131k\u00e7a g\u00f6r\u00fcld\u00fc\u011f\u00fc gibi d\u00fc\u015f\u00fcnceye h\u00e2kimdirler (detayl\u0131 inceleme i\u00e7in, bk. Clore ve Huntsinger, 2007).<\/p>\n

Peki \u201cduygu\u201d tam olarak nedir? Do\u011frusunu s\u00f6ylemek gerekirse, ger\u00e7ekten bilmiyoruz veya en az\u0131ndan tam olarak uzla\u015fam\u0131yoruz (Izard, 2010). Bu durum duygunun psikolojik temellerine dair en g\u00fcncel tart\u0131\u015fmalardan da -bazen \"100 ya\u015f\u0131ndaki duygu sava\u015f\u0131\" diye de adland\u0131r\u0131lan- rahatl\u0131kla anla\u015f\u0131labilir (Lench, Bench ve Flores, 2013; Lindquist, Siegel, Quigley ve Barrett, 2013). Neyse ki, baz\u0131 konularda genel bir anla\u015fma sa\u011flanm\u0131\u015ft\u0131r. Duygular beyin- beden- \u00e7evre etkile\u015fiminden olu\u015fan kavramsal varl\u0131klard\u0131r. Fakat onlar\u0131 beyin, beden veya \u00e7evreye bakarak bulamazs\u0131n\u0131z. Tam tersine, organizma ve \u00e7evre etkile\u015fimleri, \u00e7oklu zaman \u00f6l\u00e7ekleri ve n\u00f6robiyolojik, fizyolojik ve davran\u0131\u015fsal olarak d\u0131\u015favurumcu, eylem odakl\u0131 ve bili\u015fsel\/ bili\u015f\u00fcst\u00fc\/\u00f6znel gibi \u00e7oklu d\u00fczeyler aras\u0131ndaki de\u011fi\u015fimleri tetikledi\u011finde duygular ortaya \u00e7\u0131kar (Lewis, 2005). \u201cDuygu\u201d s\u00fcregelen durumsal ba\u011flam taraf\u0131ndan ayarlanarak bu de\u011fi\u015fimlere yans\u0131t\u0131l\u0131r. Ayn\u0131 duygusal kategori (\u00f6r. endi\u015fe) tetikleyici olaya (Tracy, 2014), biyolojik\/bili\u015fsel\/bili\u015f\u00fcst\u00fc s\u00fcre\u00e7lere (Gross, 2008; Moors, 2014) ve sosyok\u00fclt\u00fcrel etkilere ba\u011fl\u0131 olarak (Mesquita ve Boiger, 2014; Parkinson, Fischer ve Manstead, 2004) farkl\u0131 \u015fekillerde a\u00e7\u0131\u011fa \u00e7\u0131kar. \u00d6rne\u011fin, belirli ko\u015fullara ba\u011fl\u0131 olarak endi\u015feye sebep olan bir olay (topluluk \u00f6n\u00fcnde konu\u015fma, s\u0131nava girme), zamansal ba\u011flam (konu\u015fmadan bir g\u00fcn ya da bir dakika \u00f6nce), n\u00f6robiyolojik sistem (s\u0131n\u0131r \u00e7izgi uyar\u0131lmas\u0131) ve sosyal ba\u011flam (i\u015f arkada\u015flar\u0131n\u0131n ya da bir yabanc\u0131n\u0131n \u00f6n\u00fcnde konu\u015fmak) endi\u015fenin farkl\u0131 \u201ck\u0131s\u0131mlar\u0131n\u0131 tetikleyecektir. Farkl\u0131l\u0131klar\u0131n ve de\u011fi\u015fkenli\u011fin bu d\u00fczeyi insanlar ve duygular\u0131 dinamik ve uyarlanabilir oldu\u011fundan beklendiktir. De\u011fi\u015fmez duygular\u0131n evrimsel de\u011feri \u00e7ok azd\u0131r.<\/p>\n

\u00d6\u011frenme analitikleri (\u00d6A) ve e\u011fitsel veri madencili\u011fi (EVM) nereye uygundur? Bir taraftan, duygular\u0131n \u00f6\u011frenmedeki merkezi rol\u00fc g\u00f6z \u00f6n\u00fcnde tutuldu\u011funda, \u00f6\u011frenmeyi duygular\u0131 dikkate almayarak analiz etme giri\u015fimleri tamamlanmam\u0131\u015f olacakt\u0131r. Di\u011fer taraftan, duygusal fenomenlerin karma\u015f\u0131kl\u0131\u011f\u0131 ve belirsizli\u011fi d\u00fc\u015f\u00fcn\u00fcld\u00fc\u011f\u00fcnde, \u00f6\u011frenme s\u0131ras\u0131nda olu\u015fan duygular\u0131 \u00d6A ve EVM y\u00f6ntemleri olmadan analiz etme giri\u015fimleri sadece s\u0131\u011f i\u00e7g\u00f6r\u00fcler getirecektir. Neyse ki, \u00f6\u011frenme \u00fcr\u00fcnleri s\u00fcre\u00e7lerindeki duygular\u0131n olu\u015f s\u0131kl\u0131\u011f\u0131 ve etkilerini \u00e7al\u0131\u015fmak i\u00e7in veriye dayal\u0131 analitik yakla\u015f\u0131m\u0131 benimseyen bir \u00e7al\u0131\u015fma alan\u0131 vard\u0131r. Bu b\u00f6l\u00fcmde ben bu disiplinler aras\u0131 ara\u015ft\u0131rma alan\u0131ndaki temel, geli\u015fmekte olan ve gelecek temalar\u0131n baz\u0131lar\u0131n\u0131 vurguluyorum.<\/p>\n

Terminolojiye bir notla ba\u015flayal\u0131m. Duygular motivasyon, tutumlar, tercihler, fizyoloji, uyar\u0131lma ve ona at\u0131fta bulunulmakta kullan\u0131lan di\u011fer yap\u0131lar k\u00fcmesiyle ili\u015fiklidir ancak e\u015f de\u011fer de\u011fildir. Duygular ayn\u0131 zamanda miza\u00e7 ve duyu\u015fsal \u00f6zelliklerden de ayr\u0131d\u0131r (Rosenberg, 1998). Duygular hislerle de ayn\u0131 de\u011fildir. A\u00e7l\u0131k bir histir ancak duygu de\u011fildir. Ac\u0131 da de\u011fil. Duygunun ne oldu\u011funa dair bir ihtilaf da vard\u0131r. \u00d6fke kesinlikle bir duygudur ancak ya \u015fa\u015fk\u0131nl\u0131k? \u015ea\u015f\u0131rman\u0131n duyu\u015fsal bile\u015fenleri vard\u0131r (\u015fa\u015f\u0131rmaya dair hisler, y\u00fcz ifadesi karakteristikleri; (D'Mello ve Graesser, 2014b), fakat onun bir duygu olup olmad\u0131\u011f\u0131na dair bir miktar tart\u0131\u015fma bulunmaktad\u0131r (Hess, 2003; Rozin ve Cohen, 2003). Dolay\u0131s\u0131yla bu b\u00f6l\u00fcm\u00fcn geri kalan\u0131nda, daha k\u0131s\u0131tlay\u0131c\u0131 bir terim olan \u201cduygu\u201d dan ziyade daha kapsay\u0131c\u0131 bir terim olan \u201cduyu\u015fsal durumu\u201d kullanaca\u011f\u0131m.<\/p>\n\n

ANA TEMALAR<\/h2>\n

\u00d6\u011frenmedeki duyu\u015fu \u00e7al\u0131\u015fma amac\u0131yla \u00d6A\/EVM metotlar\u0131n\u0131 kullan\u0131m\u0131n\u0131 vurgulamak i\u00e7in a\u015fa\u011f\u0131daki d\u00f6rt temay\u0131 se\u00e7tim. \u00dcst\u00fcnk\u00f6r\u00fc bir bi\u00e7imde bir\u00e7ok \u00e7al\u0131\u015fmay\u0131 g\u00f6zden ge\u00e7irmektense her temada bir ya da iki \u00f6rnek te\u015fkil eden \u00e7al\u0131\u015fmay\u0131 bir belirli bir d\u00fczeyde inceliyorum. Bu da bir\u00e7ok harika \u00e7al\u0131\u015fmadan bahsedilmeyece\u011fi anlam\u0131n geliyor, fakat ben bu her tema i\u00e7in \u00e7al\u0131\u015fma alan\u0131n\u0131 ara\u015ft\u0131rma i\u015fini okuyucuya b\u0131rak\u0131yorum. Ben s\u00fcreci desteklemek ad\u0131na uygun oldu\u011funda, derleme \u00e7al\u0131\u015fmalar\u0131 \u00f6nerece\u011fim.<\/p>\n\n

T\u0131klama Ak\u0131\u015f\u0131 Verisinden Duyu\u015fsal Analiz<\/h3>\n

\u00d6A\/EVM tekniklerinin en temel kullan\u0131mlar\u0131ndan bir tanesi, \u00f6\u011frenenlerin bili\u015fsel s\u00fcre\u00e7lerini anlamak i\u00e7in \u00f6\u011frenme teknolojileri ile etkile\u015fimlerden olu\u015fturulan zengin veri ak\u0131\u015f\u0131n\u0131 kullanmakt\u0131r (Corbett ve Anderson, 1995; Sinha, Jermann, Li ve Dillenbourg, 2014). A\u015fa\u011f\u0131daki \u00e7al\u0131\u015fmada da belirtildi\u011fi \u00fczere, duyu\u015f kar\u0131\u015f\u0131ma eklendi\u011finde tamamlay\u0131c\u0131 bir dizi i\u00e7g\u00f6r\u00fc elde edilebilir.<\/p>\n

Bosch ve D'Mello (bas\u0131m a\u015famas\u0131nda) \u00f6\u011frencilerin ilk programlama oturumlar\u0131 s\u00fcresince duyu\u015fsal deneyimleri \u00fczerine bir laboratuar \u00e7al\u0131\u015fmas\u0131 yapt\u0131lar. \u00c7aylak \u00f6\u011frencilerden (N=99) 25 dakikal\u0131k desteklenmi\u015f bir \u00f6\u011frenme evresi ve 10 dakikal\u0131k desteksiz bir kaybolma evresini i\u00e7eren \u00f6z y\u00f6netimli bilgisayarl\u0131 \u00f6\u011frenme ortam\u0131n\u0131 kullanarak, Python dilinde bilgisayar programlaman\u0131n temellerini \u00f6\u011frenmeleri istendi. T\u00fcm \u00f6\u011fretimsel etkinlikler (kodlama, metin okuma, kodu test etme, hatalar\u0131 alma vb.) sistem g\u00fcnl\u00fc\u011f\u00fcne kaydedildi ve \u00f6\u011frencilerin y\u00fczlerinin videolar\u0131 ve bilgisayar ekranlar\u0131 kaydedildi. \u00d6\u011frenciler \u00f6\u011frenme oturumunun hem en ard\u0131ndan bu videolar\u0131 izleme esnas\u0131nda yakla\u015f\u0131k 100 puanda (her 15 saniyede bir) ge\u00e7mi\u015fe dair duyu\u015fsal yarg\u0131 protokol\u00fc arac\u0131l\u0131\u011f\u0131yla duyu\u015fsal h\u00fck\u00fcmler verdiler (Porayska-Pomsta, Mavrikis, D'Mello, Conati ve Baker, 2013). \u0130lgilenilen duyu\u015fsal durumlar \u00f6fke, endi\u015fe, can s\u0131k\u0131nt\u0131s\u0131, merak, i\u011frenme, korku, hayal k\u0131r\u0131kl\u0131\u011f\u0131, ak\u0131\u015f\/ me\u015fguliyet, mutluluk, \u00fcz\u00fcnt\u00fc ve \u015fa\u015fk\u0131nl\u0131k idi. Sadece me\u015fguliyet, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131, can s\u0131k\u0131nt\u0131s\u0131 ve merak yeterli s\u0131kl\u0131kta olu\u015farak daha ileri analizi gerek\u00e7elendirebildi.<\/p>\n

Yazarlar etkile\u015fim olaylar\u0131n\u0131n duygusal durumlara nas\u0131l yol a\u00e7t\u0131\u011f\u0131n\u0131 ve duygusal durumlar\u0131n \u00e7e\u015fitli davran\u0131\u015flar\u0131 nas\u0131l tetikledi\u011fini incelemi\u015ftir. Yazarlar etkile\u015fim olaylar\u0131n\u0131n duyu\u015fsal durumlara nas\u0131l sebebiyet verdi\u011fini ve duyu\u015fsal durumlar\u0131n farkl\u0131 davran\u0131\u015flar\u0131 nas\u0131l tetikledi\u011fini ara\u015ft\u0131rd\u0131lar Her bir \u00f6\u011frenen i\u00e7in desteklenmi\u015f \u00f6\u011frenme evresi boyunca etkile\u015fim olaylar\u0131 (t\u0131klama ak\u0131\u015f verisi) ve duyu\u015fsal durumlar\u0131 (\u00f6z raporlar) aras\u0131na serpi\u015ftiren zaman serileri olu\u015fturdular. Zaman serisi modelleme teknikleri (D'Mello, Taylor ve Graesser, 2007), duygusal durumlar ile etkile\u015fim olaylar\u0131 aras\u0131ndaki \u00f6nemli ge\u00e7i\u015fleri tan\u0131mlamak i\u00e7in kullan\u0131lm\u0131\u015ft\u0131r.<\/p>\n

\"\"<\/p>\n<\/a> \u015eekil 10.1. Duygusal durumlar ile bilgisayar programlama bili\u015fsel destek \u00f6\u011frenme s\u0131ras\u0131nda etkile\u015fim olaylar\u0131 aras\u0131ndaki \u00f6nemli ge\u00e7i\u015fler. D\u00fcz \u00e7izgiler duygu ge\u00e7i\u015fleri de d\u00e2hil ge\u00e7i\u015fleri g\u00f6stermektedir. Kesik \u00e7izgiler duygusal durumlar\u0131 i\u00e7ermeyen ge\u00e7i\u015fler g\u00f6sterir. Problemi G\u00f6ster: yeni bir al\u0131\u015ft\u0131rmaya ba\u015flama; Okuma: y\u00f6nergeleri ve\/ veya problem durumunu g\u00f6rme; Kodlama: mevcut kodu d\u00fczenleme veya g\u00f6rme; \u0130pucunuG\u00f6ster: ipucunu g\u00f6rme; TestRunError: kod \u00e7al\u0131\u015ft\u0131r\u0131lm\u0131\u015f ve s\u00f6z dizimi ya da \u00e7al\u0131\u015fma zaman\u0131 hatas\u0131 ile kar\u015f\u0131la\u015f\u0131lm\u0131\u015f; TestRunSuccess: kod s\u00f6z dizimi hatas\u0131 ya da \u00e7al\u0131\u015fma zaman\u0131 hatas\u0131 olmadan \u00e7al\u0131\u015ft\u0131r\u0131lm\u0131\u015f (ancak do\u011fruluk i\u00e7in kontrol yap\u0131lmam\u0131\u015f); G\u00f6nderiHatas\u0131: kod g\u00f6nderildi ve bir hata veya yanl\u0131\u015f bir cevap \u00fcretti; G\u00f6nderiBa\u015far\u0131l\u0131: kod g\u00f6nderildi ve do\u011fruydu.<\/i><\/span>\n

Sonu\u00e7ta olu\u015fan y\u00f6nl\u00fc \u00e7izge \u015eekil 10.1 de verilmi\u015ftir. Duyu\u015fsal durumlar\u0131 i\u00e7ermeyen etkile\u015fim olaylar\u0131 aras\u0131nda baz\u0131 ge\u00e7i\u015fler bulunmaktayd\u0131 (kesikli \u00e7izgiler). Bu di\u011fer etkile\u015fim olaylar\u0131na nazaran (saniyede 1 s\u0131kl\u0131kta) duygu \u00f6rneklem al\u0131m\u0131n\u0131n az s\u0131kl\u0131kta olu\u015funa (her 15 saniyede bir) ba\u011fl\u0131yd\u0131.<\/p>\n

Daha ilgi \u00e7ekici ge\u00e7i\u015fler duyu\u015fsal durumlar\u0131 i\u00e7ermekteydi. \u00d6zellikle, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131ndan \u00f6nce bir yanl\u0131\u015f bir cevap g\u00f6nderme (G\u00f6nderiHatas\u0131) gelmekte; daha sonra bu duyu\u015fsal durumlar\u0131 bir ipucu talebi (\u0130pucunuG\u00f6ster) veya kodu yap\u0131land\u0131rma (Kodlama) talebi izlemekteydi ki bu durumlar\u0131n kendisi de bir kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131 tetiklemekteydi. \u00d6\u011fretim metinlerinin okunmas\u0131 (problem tan\u0131mlamalar\u0131 d\u00e2hil) kat\u0131l\u0131m, merak, can s\u0131k\u0131nt\u0131s\u0131 ve kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n da belirtisiydi ancak hayal k\u0131r\u0131kl\u0131\u011f\u0131na u\u011framad\u0131. Bir ba\u015fka deyi\u015fle, t\u00fcm anahtar duyu\u015fsal durumlar bilgi \u00f6z\u00fcmseme (okuma) ve yap\u0131land\u0131rma (kodlama) etkinlikleriyle ili\u015fkiliydi. Ancak sadece b\u00fcy\u00fck ihtimalle \u00f6\u011frenme f\u0131rsatlar\u0131 olan kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131 ba\u015far\u0131s\u0131zl\u0131\u011fa (G\u00f6nderiHatas\u0131) ve sonras\u0131nda gelen yard\u0131m isteme davran\u0131\u015flar\u0131na (\u0130pucunu G\u00f6ster) e\u015flik etti. Birlikte bak\u0131ld\u0131\u011f\u0131nda, ge\u00e7i\u015f modeli g\u00fc\u00e7 durumlar ve sonu\u00e7ta olu\u015fan kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 veya hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131 etkinle\u015ftirici olumsuz durumlar\u0131n \u00f6\u011frenmedeki \u00f6nemli rol\u00fcn\u00fc vurgulamaktad\u0131r. (D'Mello ve Graesser, 2012b; VanLehn, Siler, Murray, Yamauchi ve Baggett, 2003). Ayn\u0131 zamanda, duygular\u0131n \u00f6\u011frenme s\u00fcrecinin i\u00e7erisine nas\u0131l serpi\u015ftirilmi\u015f oldu\u011funu a\u00e7\u0131klamaktad\u0131r.<\/p>\n\n

Etkile\u015fim \u015eemalar\u0131ndan Duygu Saptama<\/h3>\n

Duyu\u015fsal durumlar kavramsal varl\u0131klar (yap\u0131lar) oldu\u011fundan do\u011frudan do\u011fruya \u00f6l\u00e7\u00fclemezler. Ancak \u00e7evre-ki\u015fi etkile\u015fimlerinden do\u011farlar ve bili\u015fin de\u011fi\u015ftirilmesiyle eylemi etkilerler. Bu nedenle, g\u00f6zler \u00f6n\u00fcne serilen ba\u011flam ve \u00f6\u011frenen eylemleri analiz edilerek duyguyu \"\u00e7\u0131karsamak\" m\u00fcmk\u00fcn olabilmelidir. Etkile\u015fim temelli\", \"kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyas\u0131 tabanl\u0131\" veya \"alg\u0131lay\u0131c\u0131s\u0131z\" olarak adland\u0131r\u0131lan duygu saptama \u00e7al\u0131\u015fma alan\u0131 on y\u0131ldan uzun bir s\u00fcre \u00f6nce ba\u015flam\u0131\u015ft\u0131r. (Ai vd., 2006; D'Mello, Craig, Sullins ve Graesser, 2006) ve yak\u0131n zamanda Baker ve Ocumpaugh (2015) taraf\u0131ndan derlenmi\u015ftir.<\/p>\n

\u00d6rnek olarak, ortaokul -ve lise- matemati\u011fi i\u00e7in bir Ak\u0131ll\u0131 \u00d6\u011fretim Sistem (A\u00d6S) olan ve ABD'de ola\u011fan matematik \u00f6\u011fretiminin bir par\u00e7as\u0131 olarak yakla\u015f\u0131k 50, 0000 \u00f6\u011frenci taraf\u0131ndan kullan\u0131lan, ASSISTments'a duygu saptay\u0131c\u0131lar geli\u015ftiren Pardos, Baker, San Pedro ve Gowda (2013) d\u00fc\u015f\u00fcn\u00fclebilir. (Razzaq vd., 2005). Yazarlar otomatikle\u015ftirilmi\u015f duygu saptay\u0131c\u0131lar1<\/sup><\/a> olu\u015fturmak i\u00e7in denetimli bir \u00f6\u011frenme yakla\u015f\u0131m\u0131 benimsediler. Okul bilgisayar laboratuvarlar\u0131nda ASSISTments kullanan 229 \u00f6\u011frenciden e\u011fitim verisi toplad\u0131lar. \u0130nsan g\u00f6zlemciler Baker-Rodrigo G\u00f6zlem Metodu Protokol\u00fcn\u00fc kullanarak (BRGYP) \u00f6\u011frencilerin ASSISTments ile etkile\u015fimlerindeki \u00e7evrimi\u00e7i duygu g\u00f6zlemlerini (k\u0131sa ek a\u00e7\u0131klamalar) getirdiler (Ocumpaugh, Baker ve Rodrigo, 2012). Bu protokole g\u00f6re e\u011fitim alm\u0131\u015f g\u00f6zlemciler, g\u00f6zlemlenebilir davran\u0131\u015f temel al\u0131narak aray\u00fcze y\u00f6nelik a\u00e7\u0131k eylemler, akranlar ve \u00f6\u011fretmenle etkile\u015fim, beden, hareketler, el hareketleri ve y\u00fcz ifadeleri de d\u00e2hil olmak \u00fczere duyguya dair ger\u00e7ek zamanl\u0131 ek a\u00e7\u0131klamalar sa\u011flarlar. G\u00f6zlemciler d\u00f6rt duyu\u015fsal durum (can s\u0131k\u0131nt\u0131s\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131 ba\u011flanm\u0131\u015f konsantrasyon ve kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131) ve iki davran\u0131\u015f (etkinlikle ilgisiz \u015feyler yapmak veya sistemi oyunla\u015ft\u0131rmak) kodlad\u0131lar. ASSISTments kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyalar\u0131ndan \u00e7\u0131kart\u0131lan baz\u0131 \u00f6zellikler kullan\u0131larak her bir duyu\u015fsal durumu di\u011fer durumlardan ay\u0131racak (\u00f6r. can s\u0131k\u0131nt\u0131s\u0131na kar\u015f\u0131 di\u011ferleri) denetimli \u00f6\u011frenme teknikleri kullan\u0131ld\u0131. Etki tespiti do\u011frulu\u011fu, etki i\u00e7in \".632\" ile \".678\" [A-\u00fcss\u00fc metrik olarak \u00f6l\u00e7\u00fclm\u00fc\u015ft\u00fcr (EAA veya A\u0130KAA - al\u0131c\u0131 i\u015flem karakteristik e\u011frisi alt\u0131ndaki alana benzer)]- ve davran\u0131\u015flar i\u00e7in \".802\" ila \".819\" aras\u0131nda de\u011fi\u015fmi\u015ftir. S\u0131n\u0131fland\u0131r\u0131c\u0131n\u0131n do\u011frulanmas\u0131 e\u011fitim ve s\u0131nav verisi aras\u0131nda tam bir ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131 mecbur k\u0131larak, ayn\u0131 pop\u00fclasyondan yeni \u00f6\u011frencilere genellenebilirli\u011fi sa\u011flama \u015feklinde yap\u0131ld\u0131.<\/p>\n

Pardos vd. (2013) ayn\u0131 zamanda saptay\u0131c\u0131lar\u0131n\u0131n \u00f6ng\u00f6r\u00fcsel ge\u00e7erlili\u011fine dair ba\u015flang\u0131\u00e7 niteli\u011finde kan\u0131tlar sundular. Bu \u00f6l\u00e7e\u011fin geli\u015ftirilmesinden bir ka\u00e7 y\u0131l \u00f6nce 2004-2006 y\u0131llar\u0131 aras\u0131nda ASSISTments ile etkile\u015fimi olan 1393 adet farkl\u0131 bir grup \u00f6\u011frencini dosyalar\u0131na alg\u0131lay\u0131c\u0131lar\u0131n uygulanmas\u0131 ile yap\u0131lm\u0131\u015ft\u0131r. Standart hale gelmi\u015f s\u0131nav puanlar\u0131 ile otomatik olarak \u00f6l\u00e7\u00fclen duygu ve davran\u0131\u015flar k\u0131smen ili\u015fkili bulunmu\u015ftur.<\/p>\n

Sonras\u0131nda, San Pedro, Baker, Bowers ve Heffernan (2013) otomatik saptay\u0131c\u0131lara dayanarak \u00fcniversite kay\u0131tlar\u0131n\u0131 tahmin etme giri\u015fiminde bulundular. 2004'ten 2009'a kadar ASSISTments ile etkile\u015fime giren 3707 \u00f6\u011frencini mevcut kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyalar\u0131na saptay\u0131c\u0131lar\u0131 uygulad\u0131lar. Bu \u00f6\u011frencilerin \u00fcniversite kay\u0131t bilgileri Ulusal \u00d6\u011frenci B\u00fcrosundan elde edildi. Y\u0131llar sonra, olduk\u00e7a etkileyici bir bulgu olarak otomatik olarak \u00f6l\u00e7\u00fclen duyu\u015fsal durumlar\u0131n \u00fcniversiteye kay\u0131t yapt\u0131rman\u0131n en manidar \u00f6ng\u00f6r\u00fcc\u00fclerinden biri oldu\u011fu ortaya konmu\u015ftur.<\/p>\n\n

Bedensel \u0130\u015faretlerden Duygu Saptama<\/h3>\n

Duygu eylem i\u00e7in bedensel cevap sistemlerini harekete ge\u00e7irdi\u011fi i\u00e7in \u015fekillendirici bir fenomendir. Bu \u00f6\u011frenen duygusunun makine taraf\u0131ndan okunabilir bedensel sinyallere dayanarak \u00e7\u0131karsanaca\u011f\u0131n\u0131 m\u00fcmk\u00fcn k\u0131lmal\u0131d\u0131r. Bir\u00e7ok derlemde de tart\u0131\u015f\u0131ld\u0131\u011f\u0131 gibi, bedensel i\u015fretlerin duygular\u0131 saptamada kullan\u0131m\u0131 \u00fczerine zengin bir \u00e7al\u0131\u015fma alan\u0131 bulunmaktad\u0131r (Calvo ve D'Mello, 2010; D'Mello ve Kory, 2015; Zeng, Pantic, Roisman ve Huang, 2009). Ara\u015ft\u0131rma ge\u00e7mi\u015fte kontroll\u00fc ortamlardaki etkile\u015fimlere odaklanmas\u0131na kar\u015f\u0131n, ara\u015ft\u0131rmac\u0131lar bu \u00e7al\u0131\u015fmay\u0131 bilgisayar destekli s\u0131n\u0131flar ba\u015fta olmak \u00fczere ger\u00e7ek d\u00fcnyaya ta\u015f\u0131maya ba\u015flad\u0131lar. A\u015fa\u011f\u0131daki g\u00f6zden ge\u00e7irilen \u00e7al\u0131\u015fma bizim ara\u015ft\u0131rma grubumuz ve i\u015fbirlikli \u00e7al\u0131\u015ft\u0131klar\u0131m\u0131z taraf\u0131ndan ortaya konan benzer bir eme\u011fi yans\u0131tmakta ancak okuyucu Arroyo vd. (2009) bilgisayar destekli s\u0131n\u0131flarda duygu saptamaya y\u00f6nelik \u00f6nc\u00fc niteli\u011findeki \u00e7al\u0131\u015fmaya y\u00f6nlendirilmi\u015ftir.<\/p>\n

Bosch, D'Mello, Baker, Ocumpaugh ve Shute (2016) bilgisayar destekli bir s\u0131n\u0131f\u0131n karma\u015f\u0131k ger\u00e7ek d\u00fcnyas\u0131nda y\u00fcze dair \u00f6zelliklerden duygunun otomatik saptanmas\u0131n\u0131 \u00e7al\u0131\u015fm\u0131\u015flard\u0131r. Bu \u00e7al\u0131\u015fmada, 137 orta\u00f6\u011fretim ve lise \u00f6\u011frencisi k\u00fc\u00e7\u00fck gruplar halinde her zamanki fizik \/ fizik bilimleri derslerinin bir par\u00e7as\u0131 olarak iki g\u00fcn boyunca 1.5-2 saat Fizik Oyun Bah\u00e7esi (Shute, Ventura ve Kim, 2013) adl\u0131 kavramsal bir fizik e\u011fitim oyunu oynad\u0131. E\u011fitimli g\u00f6zlemciler yukar\u0131da anlat\u0131lan ASSISSTments \u00e7al\u0131\u015fmas\u0131nda oldu\u011fu gibi BRGYP saha g\u00f6zlem protokol\u00fcn\u00fc kullanarak can s\u0131k\u0131nt\u0131s\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131, ba\u011flant\u0131l\u0131- konsantrasyon ve hazza dair canl\u0131 k\u0131sa ek a\u00e7\u0131klamalar\u0131 i\u015flediler (Pardos vd., 2013). G\u00f6zlemciler ayn\u0131 zamanda \u00f6\u011frenciler konuyla alakas\u0131z olduklar\u0131nda da not ald\u0131lar.<\/p>\n

\"\"<\/p>\n<\/a> \u015eekil 10.2. Bilgisayar \u0130fade Tan\u0131ma Alet \u00e7antas\u0131 kullan\u0131larak y\u00fcz niteliklerinin otomatik taranmas\u0131. Sa\u011fdaki grafikler farkl\u0131 y\u00fcz niteliklerinin (\u00f6r. indirilmi\u015f ka\u015flar, gerilmi\u015f g\u00f6z kapaklar\u0131) muhtemel etkinle\u015ftirilmesini g\u00f6stermektedir.<\/i><\/span>\n

Oyun s\u0131ras\u0131nda \u00f6\u011frencilerin y\u00fczlerinin ve v\u00fccutlar\u0131n\u0131n \u00fcst k\u0131sm\u0131 videolar\u0131 kaydedildi ve etki a\u00e7\u0131klamalar\u0131yla senkronize edildi. Videolar, 19 y\u00fcz eylem biriminin (Ekman ve Friesen, 1978) (\u00f6r. y\u00fckseltilmi\u015f ka\u015f, gerilmi\u015f dudaklar), ba\u015f pozunun (uyumu) ve ba\u015f pozisyonunun olas\u0131l\u0131\u011f\u0131n\u0131n tahminini sa\u011flayan FACET bilgisayarl\u0131 g\u00f6r\u00fc\u015f program\u0131 (Emotient, 2014) kullan\u0131larak i\u015flendi (ekran g\u00f6r\u00fcnt\u00fcs\u00fc i\u00e7in bk. \u015eekil 10.2). Beden hareketleri hareket filtreleme algoritmalar\u0131 kullanarak tahmin edildi (Kory, D'Mello ve Olney, 2015) (bk. Fig\u00fcr 10.3). Hem y\u00fcz ifadeleri hem de bedensel hareketler kullan\u0131larak her bir duyu\u015fsal durumun saptay\u0131c\u0131lar\u0131 denetimli \u00f6\u011frenme y\u00f6ntemleri ile geli\u015ftirildi (\u00f6r. can s\u0131k\u0131nt\u0131s\u0131na kar\u015f\u0131 di\u011fer durumlar). Alg\u0131lay\u0131c\u0131lar duygu i\u00e7in .610'dan .867 aral\u0131\u011f\u0131 aras\u0131nda etkinlik d\u0131\u015f\u0131 davran\u0131\u015flar i\u00e7in ise .816 do\u011fruluklarla (miktarlar yukar\u0131da belirtilen EAA metrikleriyle \u00f6l\u00e7\u00fclerek) k\u0131smen ba\u015far\u0131l\u0131 oldular. Takip eden analizler duygu alg\u0131lay\u0131c\u0131lar\u0131n\u0131n \u00f6\u011frenciler, farkl\u0131 g\u00fcnler ve farkl\u0131 cinsiyet ve etnisiteler (insanlar taraf\u0131ndan alg\u0131land\u0131\u011f\u0131 \u015fekliyle) boyunca genelleme yapt\u0131\u011f\u0131n\u0131 do\u011frulad\u0131.<\/p>\n

\"\"<\/p>\n<\/a> \u015eekil 10.3. Beden hareketlerinin hareket sil\u00fcetleri kullan\u0131larak videodan otomatik taranmas\u0131. Sa\u011fdaki g\u00f6rsel solda oynamakta olan videodan hareketli alanlar\u0131 g\u00f6stermektedir. Alttaki grafik zaman i\u00e7indeki hareket miktar\u0131n\u0131 g\u00f6sterir.<\/i><\/span>\n

Y\u00fcz temelli duygu alg\u0131lay\u0131c\u0131lar ile ilgili bir s\u0131n\u0131rl\u0131l\u0131k, yaln\u0131zca y\u00fcz video ak\u0131\u015f\u0131nda otomatik olarak saptanabildi\u011finde uygulanabilir olmalar\u0131d\u0131r. Bu a\u015f\u0131r\u0131 hareket, kapanma, az \u0131\u015f\u0131k ve di\u011fer fakt\u00f6rlere ba\u011fl\u0131 olarak her zaman m\u00fcmk\u00fcn olmamaktad\u0131r. Asl\u0131nda y\u00fcz-temelli duygu saptay\u0131c\u0131lar t\u00fcm durumlar\u0131n %65'ine uygulanabilmektedir. Buna de\u011finmek i\u00e7in Bosch, Chen, Baker, Shute ve D'Mello (2015) \u00e7okbi\u00e7imli teknikleri etkile\u015fim temelli (\u00f6nceki b\u00f6l\u00fcme benzer olarak) ve y\u00fcz-temelli saptay\u0131c\u0131larla birle\u015ftirmek i\u00e7in kulland\u0131lar. Etkile\u015fim tabanl\u0131 saptay\u0131c\u0131lar, y\u00fcz bazl\u0131 saptay\u0131c\u0131lardan daha az do\u011fruydu (Kai vd., 2015) ancak neredeyse t\u00fcm vakalara uygulanabilirdi. Bu ikisinin birle\u015fimi ile saptay\u0131c\u0131lar\u0131n durumlara uygulanabilirli\u011fi y\u00fcz temelli saptay\u0131c\u0131lara k\u0131yasla kesinlikte k\u00fc\u00e7\u00fck bir azalma ile (<%5 farkla) %98'e y\u00fckseltilmi\u015f oldu.<\/p>\n\n

Duygu Modellerini Duygu-Bilir \u00d6\u011frenme Teknolojilerine Entegre Etmek<\/h3>\n

Yukar\u0131da tart\u0131\u015f\u0131lan etkile\u015fim ve bedensel temelli etki alg\u0131lay\u0131c\u0131lar\u0131, bir \u00f6\u011frenme teknolojisi ile etkile\u015fimler s\u0131ras\u0131nda \u00f6\u011frenenin etkilerinin ger\u00e7ek zamanl\u0131 de\u011ferlendirmelerini sa\u011flamak i\u00e7in kullan\u0131labilecek somut eserlerdir. Bu durum hissedilen duyguya dinamik anlamda cevap vererek heyecan verici bir d\u00f6ng\u00fcy\u00fc kapatma olas\u0131l\u0131\u011f\u0131n\u0131 destekler. Bu gibi duygu-bilir \u00f6\u011frenme teknolojilerinin amac\u0131, \u00f6\u011frencilerin ne hissettiklerine ek olarak ne d\u00fc\u015f\u00fcnd\u00fckleri ya da yapt\u0131klar\u0131na cevap vererek g\u00fcncel \u00f6\u011frenme teknolojilerinin uyarlanabilirlik bant aral\u0131\u011f\u0131n\u0131n geni\u015fletmektir. (\u0130nceleme i\u00e7in, bk. D\u2019Mello, Blanchard, Baker, Ocumpaugh ve Brawner, 2014). Burada ben, b\u00f6ylesine iki sisteme, Affective AutoTutor (D'Mello ve Graesser, 2012a) ve UNC-ITSPOKE (Forbes-Riley ve Litman, 2011) dikkat \u00e7ekiyorum.<\/p>\n

\"\"<\/p>\n

\u015eekil 10.4. Affective AutoTutor: \u00d6\u011frencilerin s\u0131k\u0131nt\u0131lar\u0131n\u0131, kafa kar\u0131\u015f\u0131kl\u0131klar\u0131n\u0131 ve s\u0131k\u0131nt\u0131lar\u0131n\u0131 otomatik olarak alg\u0131layan ve bunlara yan\u0131t veren konu\u015fma diyaloglar\u0131na sahip ak\u0131ll\u0131 \u00f6\u011fretim sistemi (A\u00d6S).<\/i><\/span><\/p>\n

Affective AutoTutor (bk. \u015eekil 10.4) do\u011fal dil i\u00e7inde karma-inisiyatif bir diyalog d\u00fczenleyerek \u00f6\u011frencilerin Newton fizi\u011fi, bilgisayar okur yazarl\u0131\u011f\u0131 ve bilimsel ak\u0131l y\u00fcr\u00fctme gibi zor konularda uzmanl\u0131k geli\u015ftirmelerine yard\u0131m eden konu\u015fma tabanl\u0131 bir A\u00d6S olan AutoTutor'un de\u011fi\u015ftirilmi\u015f bir s\u00fcr\u00fcm\u00fcd\u00fcr. (Graesser, Chipman, Haynes ve Olney, 2005 ). \u00d6zg\u00fcn AutoTutor sistemi \u00f6\u011frenenin bili\u015fsel durumlar\u0131na kar\u015f\u0131 duyarl\u0131 olan bir dizi belirsiz \u00fcretim kurallar\u0131na sahipti. Affective AutoTutor bu kurallar\u0131 \u00f6\u011frenenlerin duyu\u015fsal durumlar\u0131 \u00f6zellikle can s\u0131k\u0131nt\u0131s\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131n dinamik olarak de\u011ferlendirilmesine duyarl\u0131 olacak \u015fekilde artt\u0131r\u0131r. Duyu\u015fsal durumlar etkile\u015fim \u00f6r\u00fcnt\u00fclerin, b\u00fcy\u00fck beden hareketlerini ve y\u00fcz niteliklerini otomatik olarak izleyerek hissedilir. (D'Mello ve Graesser, 2012a). Affective AutoTutor duygusal g\u00f6r\u00fcn\u00fcmlerin yan\u0131 s\u0131ra empatik, y\u00fcreklendirici ve motive edici diyalog-hareketleri ile cevap verir. \u00d6rne\u011fin, asistan hafif can s\u0131k\u0131nt\u0131s\u0131na, \"Bunlar bazen biraz s\u0131k\u0131c\u0131 olabiliyor, bu y\u00fczden senin bu i\u015fin \u00fcstesinden gelmene yard\u0131mc\u0131 olmay\u0131 deneyece\u011fim. Hadi gidelim\". Duyu\u015fsal cevaplara uygun duygusal y\u00fcz ifadeleri ve duygusal olarak d\u00fczenlenmi\u015f konu\u015fmalar (\u00f6r. sentezlenmi\u015f duyguda\u015fl\u0131k veya y\u00fcreklendirme) e\u015flik eder.<\/p>\n

Affective AutoTutor'un \u00f6zg\u00fcn duyu\u015fsal olmayan AutoTutora g\u00f6re etkilili\u011fi 84 \u00f6\u011frenenin rastgele olarak iki\u015fer 30 dakikal\u0131k \u00f6\u011frenme oturumuna atand\u0131\u011f\u0131 denekler aras\u0131 desenle test edilmi\u015ftir. (D'Mello, Lehman, Sullins vd., 2010). Sonu\u00e7lar duyu\u015fsal asistan\u0131n ikinci 30 dakikal\u0131k \u00f6\u011frenme oturumunda d\u00fc\u015f\u00fck bilgi alan\u0131 \u00f6\u011frenenlerinin \u00f6\u011frenmesinde yard\u0131mc\u0131 oldu\u011funu g\u00f6stermi\u015ftir. Duyu\u015fsal asistan ilk 30 dakikal\u0131k oturumda \u00fcst bilgi alan\u0131 \u00f6\u011frenenlerinin \u00f6\u011frenmesini desteklemekte daha az etkili olmu\u015ftur. \u00d6nemli bi\u00e7imde, duyu\u015fsal asistanla \u00f6\u011frenme kazan\u0131mlar\u0131 Oturum 1'den Oturum 2'ye y\u00fckselmi\u015f oysaki duyu\u015fsal olmayan asistanla y\u00fckseli\u015f sonras\u0131 dura\u011fan bir noktaya gelmi\u015ftir. Duyu\u015fsal asistanla etkile\u015fim kuran \u00f6\u011frenenler bir sonraki transfer s\u0131nav\u0131nda \u00e7ok daha iyi bir performans g\u00f6sterdiler. Takip eden analiz g\u00f6sterdi ki, \u00f6\u011frenenlerin \u00f6\u011frenme oturumlar\u0131 boyunca bilgisayar \u00f6\u011fretenlerin (asistanlar\u0131n) insan \u00f6\u011fretenlere ne kadar benzedi\u011fine dair alg\u0131lar\u0131ndaki art\u0131\u015f asistan geri bildirim niteli\u011fine ba\u011fl\u0131d\u0131r ve \u00f6\u011frenmenin g\u00fc\u00e7l\u00fc bir yorday\u0131c\u0131s\u0131d\u0131r. (D'Mello ve Graesser, 2012c). Duyu\u015fsal asistan i\u00e7in alg\u0131daki olumlu de\u011fi\u015fim daha fazla idi.<\/p>\n

\u0130kinci bir \u00f6rnek olarak, \u00f6\u011frenenlerin s\u00f6zl\u00fc cevaplar\u0131n\u0131n kesin\/kesin olmamas\u0131 ve do\u011fru\/do\u011fru olmamas\u0131n\u0131 otomatik olarak saptama ve cevap verme becerisine sahip olan konu\u015fma yetene\u011fi etkinle\u015ftirilmi\u015f bir fizik A\u00d6S'i olan UNC-ITSPKOE (Forbes-Riley ve Litman, 2011) \u00f6rne\u011fini inceleyelim. Kesin olmaman\u0131n saptanmas\u0131 \u00f6\u011frenenlerin s\u00f6zl\u00fc cevaplar\u0131n\u0131n s\u00f6zc\u00fcksel ve diyalog tabanl\u0131 niteliklerinin yan\u0131 s\u0131ra akustik-prosodik niteliklerinin ay\u0131klanmas\u0131 ve analiz edilmesi ile ger\u00e7ekle\u015ftirilir. UNC-ITSPOKE \u00f6\u011frenenin cevab\u0131n\u0131n do\u011fru oldu\u011fu fakat emin olmad\u0131\u011f\u0131 durumlarda kesin olmama durumuna cevap verdi. Bu i\u00e7inden \u00e7\u0131k\u0131lmaz g\u00fc\u00e7 bir durumun g\u00f6stergesi olarak al\u0131nd\u0131 \u00e7\u00fcnk\u00fc \u00f6\u011frenen do\u011fru olmas\u0131na ra\u011fmen bilgisinin durumundan emin de\u011fildi. Mevcut cevap stratejisi belirsizli\u011fi \u00e7\u00f6zecek bir ek e\u011fitim sa\u011flayan a\u00e7\u0131klama temelli yan diyaloglar\u0131n faaliyete ge\u00e7irilmesini i\u00e7ermekteydi. Bu da ilaveten takip eden sorular\u0131 (daha zor i\u00e7erik i\u00e7in) veya basit\u00e7e do\u011fru bilginin ayr\u0131nt\u0131l\u0131 a\u00e7\u0131klamalarla (daha kolay i\u00e7erik i\u00e7in) tasdik edilmesini \u0130\u00e7erebilirdi.<\/p>\n

Forbes-Riley ve Litman (2011) kesin olmama durumunda uyarlanabilir cevaplar\u0131 alma (uyarlanabilir durum), kesin olmama durumunda hi\u00e7 cevap almama (uyarlanamayan durum) veya kesin olmama durumunda rastgele cevaplar alma (rastgele kontrol durumu) i\u00e7in rastgele atanan 72 tane \u00f6\u011frenenin \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131n\u0131 kar\u015f\u0131la\u015ft\u0131rd\u0131lar. Bu son durumda, yan diyaloglardan eklenen \u00f6\u011fretim i\u00e7eri\u011fi ilave \u00f6\u011fretimi kontrol etmek ad\u0131na rastgele bir dizi de\u011fi\u015fiklik i\u00e7in verilmi\u015fti. Bulgular uyarlanabilir durumun rastgele ve uyarlanamayan kontrol durumlar\u0131na nazaran az miktarda (fakat manidar olmayan d\u00fczeyde)ileri \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131 elde etti\u011fini g\u00f6sterdi. Bulgular \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131 ile ili\u015fkili olan\u0131n belki de kesin olmama durumunda uyarlanabilir cevaplar\u0131n varl\u0131\u011f\u0131 ya da yoklu\u011funun de\u011fil fakat uyarlanabilir cevaplar\u0131n say\u0131s\u0131n\u0131n oldu\u011funu a\u00e7\u0131\u011fa \u00e7\u0131kard\u0131.<\/p>\n\n

GEL\u0130\u015eMEKTE OLAN TEMALAR<\/h2>\n

Duygular, \u00f6\u011frenme, \u00d6A ve EVM'nin kesi\u015fimindeki ara\u015ft\u0131rmalar genel olarak bilgisayar destekli ak\u0131ll\u0131 \u00f6\u011fretim sistemleriyle bire bir \u00f6\u011frenmeye (Forbes-Riley ve Litman, 2011; Woolf vd., 2009), e\u011fitsel oyunlara (Conati ve Maclaren, 2009; Sabourin, Mott ve Lester 2011) veya okuma, yazma, metin-diyagram entegrasyonu ve problem \u00e7\u00f6zme gibi temel yeterlilikleri destekleyen aray\u00fczler (D'Mello ve Graesser, 2014a; D'Mello, Lehman ve Person, 2010; D'Mello ve Mills, 2014). Bu ana ara\u015ft\u0131rma kollar\u0131 olduk\u00e7a aktif olmas\u0131na ra\u011fmen, son g\u00fcnlerdeki ara\u015ft\u0131rmalar duygunun, \u00f6\u011frenmeyi kapsayan daha geni\u015f sosyok\u00fclt\u00fcrel ba\u011flam\u0131 daha yak\u0131ndan yans\u0131tacak \u00e7ok daha kapsay\u0131c\u0131 etkile\u015fim ba\u011flamlar\u0131 boyunca analizine odaklanm\u0131\u015ft\u0131r. Ben baz\u0131 heyecan verici geli\u015fmeleri \u00f6rneklendirmek a\u00e7\u0131\u015f\u0131ndan k\u0131saca d\u00f6rt ara\u015ft\u0131rma temas\u0131 tan\u0131mlayaca\u011f\u0131m.<\/p>\n\n

Y\u0131pranma ve Okul Terkinin Duygu Temelli Yorday\u0131c\u0131lar\u0131<\/h3>\n

Erken risk g\u00f6stergeleri ve erken m\u00fcdahale sistemleri \u00d6A ve EVM'nin \"muhte\u015fem uygulamalar\u01312<\/sup><\/a>\" d\u0131r. (Jayaprakash, Moody, Lauria, Regan ve Baron, 2014). Alandaki \u00e7o\u011fu sistemler akademik performans verisi, demografik veriler ve finansal yard\u0131m\u0131n ula\u015f\u0131labilirli\u011fi \u00fczerine odaklanm\u0131\u015ft\u0131r. Bu fakt\u00f6rler \u015f\u00fcphesiz \u00f6nemlidir fakat devreye girmesi muhtemel di\u011fer de\u011fi\u015fimli fakt\u00f6rler vard\u0131r. Bunu ak\u0131lda tutarak, Aguiar, Ambrose, Chawla, Goodrich ve Brockman (2014) bir m\u00fchendisli\u011fe giri\u015f dersini b\u0131rakma durumlar\u0131n\u0131 yordamada geleneksel akademik ve demografik niteliklerin yordama g\u00fcc\u00fcn\u00fc davran\u0131\u015fsal kat\u0131l\u0131m\u0131 g\u00f6steren niteliklerle kar\u015f\u0131la\u015ft\u0131rd\u0131lar. Temel bulgular\u0131, oturum a\u00e7ma say\u0131lar\u0131, g\u00f6nderilen \u00e7al\u0131\u015fmalar\u0131n say\u0131s\u0131 ve sayfa t\u0131klanma say\u0131lar\u0131yla \u00f6l\u00e7\u00fclen e-portfolyolarla davran\u0131\u015fsal olarak me\u015fguliyet durumunun ders b\u0131rakmay\u0131, sadece akademik performans ve demografik yap\u0131lardan olu\u015fturulan modellerden daha iyi yordayabildi\u011fiydi. Bu \u00e7al\u0131\u015fmada duygu do\u011frudan olarak \u00f6l\u00e7\u00fclmemi\u015f olsa da e portfolyolara davran\u0131\u015fsal olarak kat\u0131l\u0131m, g\u00fc\u00e7l\u00fc bir g\u00fcd\u00fcleyici duygu olan ilginin bir i\u015fareti olarak d\u00fc\u015f\u00fcn\u00fclebilir.<\/p>\n\n

Tart\u0131\u015fma Forumlar\u0131n\u0131n Duygu Analizleri<\/h3>\n

Dil duygular\u0131 ileti\u015fime ge\u00e7irir. Dolay\u0131s\u0131yla duygu analizi ve fikir madencili\u011fi teknikleri (Pang ve Lee, 2008) \u00f6\u011frencileri bir \u00f6\u011frenme deneyimi hakk\u0131ndaki d\u00fc\u015f\u00fcncelerinin (yaz\u0131l\u0131 dilde ifade edilen) ili\u015fkili davran\u0131\u015flar\u0131 (\u00f6zellikle y\u0131pranma) nas\u0131l yordad\u0131\u011f\u0131n\u0131 \u00e7al\u0131\u015f\u0131rken \u00f6nemli d\u00fczeyde bir g\u00fcce sahiptir. Bu do\u011frultuda Wen, Yang ve Rose (2014) \u00fc\u00e7 tane kitlesel a\u00e7\u0131k \u00e7evrimi\u00e7i derslerdeki (KA\u00c7D) \u00f6\u011frenen g\u00f6nderilerine duygu analizi tekniklerini uygulad\u0131lar. Olumlunun olumsuz terimlere oran\u0131 ile okul b\u0131rakman\u0131n zamana oran\u0131 aras\u0131nda negatif ili\u015fki g\u00f6zlemlediler. Daha da yak\u0131n d\u00f6nemde, Yang, Wen, Howley, Kraut ve Rose (2015) \u00f6\u011frencideki kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n belirleyicisi olan tart\u0131\u015fma g\u00f6nderilerini otomatik olarak belirleyecek metotlar geli\u015ftirdiler. Kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n okulu s\u00fcrd\u00fcrebilme olas\u0131l\u0131\u011f\u0131n\u0131 d\u00fc\u015f\u00fcrd\u00fc\u011f\u00fcn\u00fc, fakat bunun kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131 \u00e7\u00f6zme ve di\u011fer destekleyici m\u00fcdahalelerle hafifletilebilece\u011fini g\u00f6sterdiler.<\/p>\n\n

S\u0131n\u0131f \u00d6\u011frenme Analitikleri<\/h3>\n

Alg\u0131lama ve sinyal i\u015fleme teknolojilerindeki son geli\u015fmeler daha \u00f6nceden sadece ki\u015fisel raporlar ve kullan\u0131\u015fs\u0131z insan g\u00f6zlemleri ile elde edilebilecek olan \u00f6\u011frencilerin s\u0131n\u0131f deneyimlerine ili\u015fkin halleri otomatik olarak modelleyebilmeyi sa\u011flad\u0131. \u00d6rne\u011fin, ikinci ku\u015fak Kinects g\u00f6zlerin ya da a\u011fz\u0131n a\u00e7\u0131k olup olmad\u0131\u011f\u0131n\u0131, ki\u015finin ba\u015fka bir yere bak\u0131p bakmad\u0131\u011f\u0131n\u0131 ya da a\u011fz\u0131n\u0131n oynay\u0131p oynamad\u0131\u011f\u0131n\u0131 bir defada alt\u0131 ki\u015fiye kadar tespit edebilmektedir. (Microsoft, 2015). \u00d6nc\u00fc bir \u00e7al\u0131\u015fmada, Raca, Kidzinski ve Dillenbourg (2015), \u00f6\u011frencileri karatahta alan\u0131n\u0131n etraf\u0131na yerle\u015ftirilmi\u015f birden fazla kamera kullanarak s\u0131n\u0131fta izlemi\u015ftir. Daha sonra bir saptay\u0131c\u0131y\u0131 e\u011fitmek i\u00e7in kullan\u0131lan kafa alg\u0131lama ve kafa pozu tahmini i\u00e7in bilgisayarl\u0131 g\u00f6rme teknikleri kullan\u0131lm\u0131\u015ft\u0131r. \u00d6\u011frencilerin dikkatini \u00e7ekme (kendi kendine raporlama yoluyla do\u011fruland\u0131). \u00c7ok modlu \u00f6\u011frenme analiti\u011fi alan\u0131yla (Blikstein, 2013) ili\u015fkili olan geli\u015fmekteki bu alan, \u00f6n\u00fcm\u00fczdeki uzun y\u0131llar i\u00e7inde \u00f6nemli geli\u015fmelere haz\u0131rd\u0131r.<\/p>\n\n

\u00d6\u011fretmen Analitikleri<\/h3>\n

\u00d6\u011fretmen uygulamalar\u0131n\u0131n \u00f6\u011frenci duygu ve kat\u0131l\u0131mlar\u0131n\u0131 etkiledi\u011fi bilindi\u011finden \u00f6\u011fretmenler d\u00f6ng\u00fcn\u00fcn d\u0131\u015f\u0131nda b\u0131rak\u0131lmamal\u0131d\u0131r. Ne yaz\u0131k ki, \u00f6\u011fretmenlerin \u00f6\u011fretim uygulamalar\u0131n\u0131n niceli\u011fi s\u0131n\u0131flardaki canl\u0131 g\u00f6zlemlere dayan\u0131r (\u00f6r. Nystrand, 1997) ve bu da ara\u015ft\u0131rman\u0131n \u00f6l\u00e7eklendirilmesini zorla\u015ft\u0131r\u0131r. Bunu ele almak i\u00e7in, ara\u015ft\u0131rmac\u0131lar \u00f6\u011fretmen \u00f6\u011fretimsel uygulamalar\u0131n\u0131n otomatik analizi i\u00e7in y\u00f6ntemler geli\u015ftirmeye ba\u015flad\u0131lar. \u00d6nc\u00fc olan \u00e7al\u0131\u015fmalardan birinde, Wang, Miller ve Cortina (2013) 1 ve 3'\u00fcnc\u00fc s\u0131n\u0131f matematik derslerinin ses kay\u0131tlar\u0131n\u0131 alm\u0131\u015flar ve bu derslerdeki tart\u0131\u015fmalar\u0131n d\u00fczeylerinin belirleyecek otomatik metotlar geli\u015ftirmi\u015flerdir. Bu \u00e7al\u0131\u015fma yak\u0131n zamanda daha b\u00fcy\u00fck orta \u00f6\u011fretim alan yaz\u0131n \u00f6rneklemeleri ve sadece \u00f6\u011fretmen sesi kullan\u0131lan dil s\u0131n\u0131f\u0131 \u00f6rneklemelerinde bir \u00e7ok ek \u00f6\u011fretim etkinli\u011fi (ders anlatma, k\u00fc\u00e7\u00fck grup \u00e7al\u0131\u015fmas\u0131, denetimli s\u0131ra \u00e7al\u0131\u015fmas\u0131, soru\/cevap ve y\u00f6nergeler ve y\u00f6nler) (Donnelly vd., 2016a) veya \u00f6\u011fretmen ve s\u0131n\u0131f ses kayd\u0131n\u0131n birle\u015ftirilmesini (Donnelly vd., 2016b) analiz etmek amac\u0131yla geni\u015fletilmi\u015ftir. Blanchard vd. (2016) insanlar taraf\u0131ndan kodlanan sorular ile .85 korelasyon sa\u011flayarak, \u00f6\u011fretmen sorular\u0131n\u0131 otomatik olarak saptayabilmek i\u00e7in \u00f6\u011fretmen ses kay\u0131tlar\u0131n\u0131 kulland\u0131. Bu alandaki \u00e7al\u0131\u015fmalarda bir sonraki a\u015fama ise, \u00f6\u011frencilerin nas\u0131l hissettiklerini bir de ne d\u00fc\u015f\u00fcnd\u00fckleri, yapt\u0131klar\u0131 ve \u00f6\u011frendiklerini etkileyen \u00e7evredeki di\u011fer \u00f6gelerle birlikte de\u011ferlendirebilmek i\u00e7in \u00f6\u011fretmenlerin ne yapt\u0131klar\u0131na dair bilgiyi kullanmak olacakt\u0131r.<\/p>\n\n

GELECEK TEMALARI<\/h2>\n

Ara\u015ft\u0131rmaya dair baz\u0131 olas\u0131 gelecek temalar\u0131n\u0131 k\u0131saca vurgulayarak bitireyim. Umut verici ara\u015ft\u0131rma alanlar\u0131ndan biri \u00f6\u011frenenlerin ve \u00f6\u011frenme topluluklar\u0131n\u0131n duygusal deneyimlerinin geleneksel bir s\u0131n\u0131f\u0131n, ters y\u00fcz edilmi\u015f s\u0131n\u0131f\u0131n, ya da KA\u00c7D '\u0131n geni\u015fletilmi\u015f zaman \u00f6l\u00e7e\u011finde detayl\u0131 bir analizini i\u00e7ermektedir. (Dillon vd., 2016). \u0130kincisi faydal\u0131 olanlar meydana \u00e7\u0131kar\u0131labilsin diye (\u00f6r. Strain ve D'Mello, 2014), \u00f6\u011frenme esnas\u0131ndaki duygu d\u00fczenlemenin, \u00f6zellikle \u00d6A\/EVM metotlar\u0131n\u0131n farkl\u0131 d\u00fczenleme stratejilerini belirlemede nas\u0131l kullan\u0131labilece\u011finin \u00e7al\u0131\u015f\u0131lmas\u0131d\u0131r (Gross, 2008). \u00dc\u00e7\u00fcnc\u00fcs\u00fc duygunun yan\u0131 s\u0131ra fark\u0131ndal\u0131k dikkat durumlar\u0131, konu ile ili\u015fkili olmayan durumlar\u0131n d\u00fc\u015f\u00fcn\u00fclmesi ve duygu- dikkat kar\u0131\u015f\u0131m\u0131n\u0131n \"ak\u0131\u015f-deneyimi\" ne benzer \u015fekilde nas\u0131l harmanland\u0131\u011f\u0131n\u0131n (Csikszentmihalyi, 1990) ortaya \u00e7\u0131k\u0131\u015f\u0131 ve beden ve davran\u0131\u015fta g\u00f6r\u00fcn\u00fcr olu\u015funu birlikte d\u00fc\u015f\u00fcnecektir. D\u00f6rd\u00fcnc\u00fcs\u00fc s\u00f6zde \"bili\u015fsel olmayan\" (Farrington vd., 2012) azim, oto kontrol ve gayret gibi ki\u015fisel \u00f6zelliklerin \u00f6\u011frenen duygular\u0131n\u0131 ve onlar\u0131 d\u00fczenleme \u00e7abalar\u0131n\u0131 nas\u0131l denetledi\u011fine de\u011finir (\u00f6r. Galla vd., 2014). Be\u015fincisi i\u015f birli\u011fi yapman\u0131n kritik bir 21. y\u00fczy\u0131l becerisi olarak \u00f6nemine istinaden (OECD, 2015), i\u015fbirlikli \u00f6\u011frenme ve problem \u00e7\u00f6zme esnas\u0131nda \u00f6\u011frenen gruplar\u0131n\u0131n duygular\u0131n\u0131 izleyebilir (Ringeval, Sonderegger, Sauer ve Lalanne, 2013).<\/p>\n

Son olarak, William James'in 1884 tarihli duygu \u00fczerine klasik tezinden al\u0131nt\u0131 yaparak \"\u00c7evremin par\u00e7alar\u0131n\u0131n en \u00f6nemlisi benim adam\u0131md\u0131r. Onun bana olan tutumunun bilinci; utan\u00e7lar\u0131m\u0131n, k\u0131zg\u0131nl\u0131klar\u0131m\u0131n ve korkular\u0131m\u0131n \u00e7o\u011funun normal olarak \u00e7\u00f6zecek olan alg\u0131d\u0131r\" (s. 195). G\u00fcn\u00fcm\u00fcze kadar olan ara\u015ft\u0131rmalar temelde ba\u015far\u0131, bilgiye dair ve konuya dair duygulara odaklanm\u0131\u015ft\u0131r. Ancak \u00f6\u011frenmenin ger\u00e7ekle\u015fti\u011fi sosyo k\u00fclt\u00fcrel ba\u011flam\u0131n analizi mutlaka gurur, su\u00e7luluk, k\u0131skan\u00e7l\u0131k, haset gibi sosyal duygulara duygun bir \u015fekilde de\u011finmelidir. Bu hem bir gelecek temas\u0131 ve hem de b\u00fcy\u00fck bir ara\u015ft\u0131rma meydan okumas\u0131d\u0131r.<\/p>\n\n

SONU\u00c7<\/h2>\n

\u00d6\u011frenme so\u011fuk bir entelekt\u00fcel etkinlik de\u011fildir; duygularla i\u015faretlenmi\u015ftir. Duygular sadece ifadelerin s\u00fcs\u00fc de\u011fildir ayn\u0131 zamanda temsil g\u00f6revleri de vard\u0131r. Fakat duygu \u00e7oklu zaman \u00f6l\u00e7eklerinde devingen olarak geli\u015fen \u00e7oklu bile\u015fenlere sahip karma\u015f\u0131k bir fenomendir. Duyu\u015fsal bilimler ve duyu\u015fsal sinir bilimdeki b\u00fcy\u00fck ad\u0131mlara ra\u011fmen duygular hakk\u0131nda \u00e7ok az ve hatta \u00f6\u011frenme esnas\u0131ndaki duygular hakk\u0131nda daha da az \u015fey biliyoruz. Bu kesinlikle teorik olarak bir belirginlik olu\u015fana kadar duygular\u0131 modellemekten ka\u00e7\u0131nmam\u0131z gerekti\u011fi anlam\u0131na gelmez. Ger\u00e7ekte durum tam tersidir. Basit\u00e7e duygular\u0131 modelledi\u011fimizi s\u00f6yledi\u011fimizde neyi modelledi\u011fimiz konusunda daha dikkatli olmal\u0131y\u0131z anlam\u0131na gelir. Ayn\u0131 zamanda duygunun i\u00e7indeki karma\u015f\u0131kl\u0131\u011f\u0131 ve belirsizli\u011fin etkisini azaltmak yerine onlar\u0131 benimsemeliyiz. Bilakis, bulgu temelli, veriye dayal\u0131, \u00d6A ve EVM'nin analitik y\u00f6ntemlerinden biri ger\u00e7ek ya\u015fam veri toplaman\u0131n beraberinde hem \u00f6\u011frenme bilimi hem de duygu bilimine geli\u015ftirecek benzersiz potansiyele sahiptir. Her \u015fey \u00f6\u011frenme analizine duygular\u0131 dahil ederek ba\u015flar.<\/p>\n\n

TE\u015eEKK\u00dcR B\u00d6L\u00dcM\u00dc<\/h2>\n

Bu ara\u015ft\u0131rma Ulusal Bilim Vakf\u0131 (UBV) (DRL 1108845 ve IIS 1523091), Bill ve Melinda Gates Vakf\u0131 ve E\u011fitim Bilimleri Enstit\u00fcs\u00fc (R305A130030) taraf\u0131ndan desteklenmi\u015ftir. Bu makalede ifade edilen d\u00fc\u015f\u00fcnceler, bulgular ve \u00e7\u0131kar\u0131mlar veya tavsiyeler yazara ait olup, destek veren kurulu\u015flar\u0131n g\u00f6r\u00fc\u015flerini yans\u0131tmas\u0131 gerekmemektedir.<\/p>\n\n

KAYNAK\u00c7A<\/h2>\nAguiar, E., Ambrose, G. A. A., Chawla, N. V., Goodrich, V., & Brockman, J. (2014). Engagement vs. performance: Using electronic portfolios to predict first semester engineering student persistence. Journal of Learning Analytics, 1<\/i>(3), 7\u201333. <\/span>\n\nAi, H., Litman, D. J., Forbes-Riley, K., Rotaru, M., Tetreault, J. R., & Pur, A. (2006). Using system and user performance features to improve emotion detection in spoken tutoring dialogs. Proceedings of the 9th International Conference on Spoken Language Processing <\/i>(Interspeech 2006). <\/span>\n\nArroyo, I., Woolf, B., Cooper, D., Burleson, W., Muldner, K., & Christopherson, R. (2009). Emotion sensors go to school. In V. Dimitrova, R. Mizoguchi, B. Du Boulay, & A. Graesser (Eds.), Proceedings of the 14th International Conference on Artificial Intelligence in Education <\/i>(pp. 17\u201324). Amsterdam: IOS Press. <\/span>\n\nBaker, R., & Ocumpaugh, J. (2015). Interaction-based affect detection in educational software. In R. Calvo, S. D'Mello, J. Gratch, & A. Kappas (Eds.), The Oxford handbook of affective computing <\/i>(pp. 233\u2013245). New York: Oxford University Press. <\/span>\n\nBarth, C. M., & Funke, J. (2010). Negative affective environments improve complex solving performance. Cognition and Emotion, 24<\/i>(7), 1259\u20131268. doi:10.1080\/02699930903223766 <\/span>\n\nBlanchard, N., Donnelly, P., Olney, A. M., Samei, B., Ward, B., Sun, X., . . . D'Mello, S. K. (2016). Identifying teacher questions using automatic speech recognition in live classrooms. Proceedings of the 17th Annual SIGdial Meeting on Discourse and Dialogue <\/i>(SIGDIAL 2016) (pp. 191\u2013201). Association for Computational Linguistics. <\/span>\n\nBlikstein, P. (2013). Multimodal learning analytics. Proceedings of the 3rd International Conference on Learning Analytics and Knowledge <\/i>(LAK\u201913), 8\u201312 April 2013, Leuven, Belgium. New York: ACM. <\/span>\n\nBosch, N., Chen, H., Baker, R., Shute, V., & D'Mello, S. K. (2015). Accuracy vs. availability heuristic in multimodal affect detection in the wild. Proceedings of the 17th ACM International Conference on Multimodal Interaction <\/i>(ICMI 2015). New York: ACM. <\/span>\n\nBosch, N., & D'Mello, S. K. (in press). The affective experience of novice computer programmers. International Journal of Artificial Intelligence in Education<\/i>. <\/span>\n\nBosch, N., D'Mello, S., Baker, R., Ocumpaugh, J., & Shute, V. (2016). Using video to automatically detect learner affect in computer-enabled classrooms. ACM Transactions on Interactive Intelligent Systems, 6<\/i>(2), doi:10.1145\/2946837. <\/span>\n\nCalvo, R. A., & D'Mello, S. K. (2010). Affect detection: An interdisciplinary review of models, methods, and their applications. IEEE Transactions on Affective Computing, 1<\/i>(1), 18\u201337. doi:10.1109\/T-AFFC.2010.1 <\/span>\n\nClore, G. L., & Huntsinger, J. R. (2007). How emotions inform judgment and regulate thought. Trends in Cognitive Sciences, 11<\/i>(9), 393\u2013399. doi: 10.1016 \/ j.tics.2007.08.005 <\/span>\n\nConati, C., & Maclaren, H. (2009). Empirically building and evaluating a probabilistic model of user affect. User Modeling and User-Adapted Interaction, 19<\/i>(3), 267\u2013303. <\/span>\n\nCorbett, A., & Anderson, J. (1995). Knowledge tracing: Modeling the acquisition of procedural knowledge. User Modeling and User-Adapted Interaction, 4<\/i>(4), 253\u2013278. <\/span>\n\nCsikszentmihalyi, M. (1990). Flow: The psychology of optimal experience<\/i>. New York: Harper & Row. <\/span>\n\nD'Mello, S. K. (2016). On the influence of an iterative affect annotation approach on inter-observer and self-observer reliability. IEEE Transactions on Affective Computing, 7<\/i>(2), 136\u2013149. <\/span>\n\nD'Mello, S. K., Blanchard, N., Baker, R., Ocumpaugh, J., & Brawner, K. (2014). I feel your pain: A selective review of affect-sensitive instructional strategies. In R. Sottilare, A. Graesser, X. Hu, & B. Goldberg (Eds.), Design recommendations for adaptive intelligent tutoring systems: Adaptive instructional strategies <\/i>(Vol. 2, pp. 35\u201348). Orlando, FL: US Army Research Laboratory. <\/span>\n\nD'Mello, S., Craig, S., Sullins, J., & Graesser, A. (2006). Predicting affective states expressed through an emote-aloud procedure from AutoTutor's mixed-initiative dialogue. International Journal of Artificial Intelligence in Education, 16<\/i>(1), 3\u201328.<\/span>\n\nD'Mello, S., & Graesser, A. (2012a). AutoTutor and Affective AutoTutor: Learning by talking with cognitively and emotionally intelligent computers that talk back. ACM Transactions on Interactive Intelligent Systems, 2<\/i>(4), 23:22\u201323:39. <\/span>\n\nD'Mello, S., & Graesser, A. (2012b). Dynamics of affective states during complex learning. Learning and Instruction, 22<\/i>(2), 145\u2013157. doi: 10.1016 \/ j.learninstruc.2011.10.001 <\/span>\n\nD'Mello, S., & Graesser, A. (2012c). Malleability of students' perceptions of an affect-sensitive tutor and its influence on learning. In G. Youngblood & P. McCarthy (Eds.), Proceedings of the 25th Florida Artificial Intelligence Research Society Conference <\/i>(pp. 432\u2013437). Menlo Park, CA: AAAI Press. <\/span>\n\nD'Mello, S., & Graesser, A. (2014a). Inducing and tracking confusion and cognitive disequilibrium with breakdown scenarios. Acta Psychologica, 151<\/i>, 106\u2013116. <\/span>\n\nD'Mello, S. K., & Graesser, A. C. (2014b). Confusion. In R. Pekrun & L. Linnenbrink-Garcia (Eds.), International handbook of emotions in education <\/i>(pp. 289\u2013310). New York: Routledge. <\/span>\n\nD'Mello, S. K., & Kory, J. (2015). A review and meta-analysis of multimodal affect detection systems. ACM Computing Surveys, 47<\/i>(3), 43:41\u201343:46. <\/span>\n\nD'Mello, S., Lehman, B., & Person, N. (2010). Monitoring affect states during effortful problem solving activities. International Journal of Artificial Intelligence in Education, 20<\/i>(4), 361\u2013389. <\/span>\n\nD'Mello, S., Lehman, B., Sullins, J., Daigle, R., Combs, R., Vogt, K., . . . Graesser, A. (2010). A time for emoting: When affect-sensitivity is and isn't effective at promoting deep learning. In J. Kay & V. Aleven (Eds.), Proceedings of the 10th International Conference on Intelligent Tutoring Systems <\/i>(pp. 245\u2013254). Berlin \/ Heidelberg: Springer. <\/span>\n\nD'Mello, S., & Mills, C. (2014). Emotions while writing about emotional and non-emotional topics. Motivation and Emotion, 38<\/i>(1), 140\u2013156. <\/span>\n\nD'Mello, S., Taylor, R. S., & Graesser, A. (2007). Monitoring affective trajectories during complex learning. In D. McNamara & J. Trafton (Eds.), Proceedings of the 29th Annual Cognitive Science Society <\/i>(pp. 203\u2013208). Austin, TX: Cognitive Science Society. <\/span>\n\nDarwin, C. (1872). The expression of the emotions in man and animals. London: John Murray. <\/span>\n\nDillon, J., Bosch, N., Chetlur, M., Wanigasekara, N., Ambrose, G. A., Sengupta, B., & D'Mello, S. K. (2016). Student emotion, co-occurrence, and dropout in a MOOC context. In T. Barnes, M. Chi, & M. Feng (Eds.), Proceedings of the 9th International Conference on Educational Data Mining <\/i>(EDM 2016) (pp. 353\u2013357). International Educational Data Mining Society. <\/span>\n\nDonnelly, P., Blanchard, N., Samei, B., Olney, A. M., Sun, X., Ward, B., . . . D'Mello, S. K. (2016a). Automatic teacher modeling from live classroom audio. In L. Aroyo, S. D'Mello, J. Vassileva, & J. Blustein (Eds.), Proceedings of the 2016 ACM International Conference on User Modeling, Adaptation, & Personalization <\/i>(ACM UMAP 2016) (pp. 45\u201353). New York: ACM. <\/span>\n\nDonnelly, P., Blanchard, N., Samei, B., Olney, A. M., Sun, X., Ward, B., . . . D'Mello, S.K. (2016b). Multi-sensor modeling of teacher instructional segments in live classrooms. Proceedings of the 18th ACM International Conference on Multimodal Interaction <\/i>(ICMI 2016). New York: ACM. <\/span>\n\nEkman, P., & Friesen, W. (1978). The facial action coding system: A technique for the measurement of facial movement. Palo Alto, CA: Consulting Psychologists Press. <\/span>\n\nEmotient. (2014). FACET: Facial expression recognition software. <\/span>\n\nFarrington, C. A., Roderick, M., Allensworth, E., Nagaoka, J., Keyes, T. S., Johnson, D. W., & Beechum, N. O. (2012). Teaching adolescents to become learners: The role of noncognitive factors in shaping school performance: A critical literature review. Chicago, IL: University of Chicago Consortium on Chicago School Research. <\/span>\n\nForbes-Riley, K., & Litman, D. J. (2011). Benefits and challenges of real-time uncertainty detection and adaptation in a spoken dialogue computer tutor. Speech Communication, 53<\/i>(9\u201310), 1115\u20131136. doi:10.1016\/j. specom.2011.02.006<\/span>\n\nFredrickson, B., & Branigan, C. (2005). Positive emotions broaden the scope of attention and thought-action repertoires. Cognition & Emotion, 19<\/i>(3), 313\u2013332. doi:10.1080\/02699930441000238 <\/span>\n\nGalla, B. M., Plummer, B. D., White, R. E., Meketon, D., D'Mello, S. K., & Duckworth, A. L. (2014). The academic diligence task (ADT): Assessing individual differences in effort on tedious but important schoolwork. Contemporary Educational Psychology, 39<\/i>(4), 314\u2013325. <\/span>\n\nGraesser, A., Chipman, P., Haynes, B., & Olney, A. (2005). AutoTutor: An intelligent tutoring system with mixed-initiative dialogue. IEEE Transactions on Education, 48<\/i>(4), 612\u2013618. doi:10.1109\/TE.2005.856149 <\/span>\n\nGross, J. (2008). Emotion regulation. In M. Lewis, J. Haviland-Jones, & L. Barrett (Eds.), Handbook of emotions <\/i>(3rd ed., pp. 497\u2013512). New York: Guilford. <\/span>\n\nHess, U. (2003). Now you see it, now you don't - the confusing case of confusion as an emotion: Commentary on Rozin and Cohen (2003). Emotion, 3<\/i>(1), 76\u201380. <\/span>\n\nIsen, A. (2008). Some ways in which positive affect influences decision making and problem solving. In M. Lewis, J. Haviland-Jones, & L. Barrett (Eds.), Handbook of emotions <\/i>(3rd ed., pp. 548\u2013573). New York: Guilford. <\/span>\n\nIzard, C. (2010). The many meanings\/aspects of emotion: Definitions, functions, activation, and regulation. Emotion Review, 2<\/i>(4), 363\u2013370. doi:10.1177\/1754073910374661 <\/span>\n\nJames, W. (1884). What is an emotion? Mind, 9<\/i>, 188\u2013205. <\/span>\n\nJayaprakash, S. M., Moody, E. W., Laur\u00eda, E. J., Regan, J. R., & Baron, J. D. (2014). Early alert of academically at-risk students: An open source analytics initiative. Journal of Learning Analytics, 1<\/i>(1), 6\u201347. <\/span>\n\nKai, S., Paquette, L., Baker, R., Bosch, N., D'Mello, S., Ocumpaugh, J., . . . Ventura, M. (2015). Comparison of face-based and interaction-based affect detectors in physics playground. In C. Romero, M. Pechenizkiy, J. Boticario, & O. Santos (Eds.), Proceedings of the 8th International Conference on Educational Data Mining <\/i>(EDM 2015) (pp. 77\u201384). International Educational Data Mining Society. <\/span>\n\nKory, J., D'Mello, S. K., & Olney, A. (2015). Motion tracker: Camera-based monitoring of bodily movements using motion silhouettes. PloS ONE, 10<\/i>(6), doi:10.1371\/journal.pone.0130293. <\/span>\n\nLench, H. C., Bench, S. W., & Flores, S. A. (2013). Searching for evidence, not a war: Reply to Lindquist, Siegel, Quigley, and Barrett (2013). Psychological Bulletin, 113<\/i>(1), 264\u2013268. <\/span>\n\nLewis, M. D. (2005). Bridging emotion theory and neurobiology through dynamic systems modeling. Behavioral and Brain Sciences, 28<\/i>(2), 169\u2013245. <\/span>\n\nLindquist, K. A., Siegel, E. H., Quigley, K. S., & Barrett, L. F. (2013). The hundred-year emotion war: Are emotions natural kinds or psychological constructions? Comment on Lench, Flores, and Bench (2011). Psychological Bulletin, 139<\/i>(1), 264\u2013268. <\/span>\n\nMcKeown, G. J., & Sneddon, I. (2014). Modeling continuous self-report measures of perceived emotion using generalized additive mixed models. Psychological Methods, 19<\/i>(1), 155\u2013174. <\/span>\n\nMesquita, B., & Boiger, M. (2014). Emotions in context: A sociodynamic model of emotions. Emotion Review, 6<\/i>(4), 298\u2013302. <\/span>\n\nMicrosoft. (2015). Kinect for Windows SDK MSDN. Retrieved 5 August 2015 from https:\/\/msdn.microsoft.com\/ en-us\/library\/dn799271.aspx <\/span>\n\nMoors, A. (2014). Flavors of appraisal theories of emotion. Emotion Review, 6<\/i>(4), 303\u2013307. <\/span>\n\nNystrand, M. (1997). Opening dialogue: Understanding the dynamics of language and learning in the English classroom. Language and Literacy Series<\/i>. New York: Teachers College Press. <\/span>\n\nOcumpaugh, J., Baker, R. S., & Rodrigo, M. M. T. (2012). Baker-Rodrigo Observation Method Protocol (BROMP) 1.0. Training Manual version 1.0. New York. <\/span>\n\nOECD (Organisation for Economic Co-operation and Development). (2015). PISA 2015 collaborative problem solving framework.<\/span>\n\nOhman, A., & Soares, J. (1994). Unconscious anxiety: Phobic responses to masked stimuli. Journal of Abnormal Psychology, 103<\/i>(2), 231\u2013240. <\/span>\n\nPang, B., & Lee, L. (2008). Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 2<\/i>(1\u20132), 1\u2013135. <\/span>\n\nPardos, Z., Baker, R. S. J. d., San Pedro, M. O. C. Z., & Gowda, S. M. (2013). Affective states and state tests: Investigating how affect throughout the school year predicts end of year learning outcomes. In D. Suthers, K. Verbert, E. Duval, & X. Ochoa (Eds.), Proceedings of the 3rd International Conference on Learning Analytics and Knowledge <\/i>(LAK\u201913), 8\u201312 April 2013, Leuven, Belgium (pp. 117\u2013124). New York: ACM. <\/span>\n\nParkinson, B., Fischer, A. H., & Manstead, A. S. (2004). Emotion in social relations: Cultural, group, and interpersonal processes<\/i>. Psychology Press. <\/span>\n\nPekrun, R., & Stephens, E. J. (2011). Academic emotions. In K. Harris, S. Graham, T. Urdan, S. Graham, J. Royer, & M. Zeidner (Eds.), APA educational psychology handbook, Vol 2: Individual differences and cultural and contextual factors <\/i>(pp. 3\u201331). Washington, DC: American Psychological Association. <\/span>\n\nPorayska-Pomsta, K., Mavrikis, M., D'Mello, S. K., Conati, C., & Baker, R. (2013). Knowledge elicitation methods for affect modelling in education. International Journal of Artificial Intelligence in Education, 22<\/i>, 107\u2013140. <\/span>\n\nRaca, M., Kidzinski, L., & Dillenbourg, P. (2015). Translating head motion into attention: Towards processing of student's body language. Proceedings of the 8th International Conference on Educational Data Mining<\/i>. International Educational Data Mining Society. <\/span>\n\nRazzaq, L., Feng, M., Nuzzo-Jones, G., Heffernan, N. T., Koedinger, K. R., Junker, B., . . . Choksey, S. (2005). The assistment project: Blending assessment and assisting. In C. Loi & G. McCalla (Eds.), Proceedings of the 12th International Conference on Artificial Intelligence in Education <\/i>(pp. 555\u2013562). Amsterdam: IOS Press. <\/span>\n\nRingeval, F., Sonderegger, A., Sauer, J., & Lalanne, D. (2013). Introducing the RECOLA multimodal corpus of remote collaborative and affective interactions. Proceedings of the 2nd International Workshop on Emotion Representation, Analysis and Synthesis in Continuous Time and Space (EmoSPACE) in conjunction with the 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition<\/i>. Washington, DC: IEEE. <\/span>\n\nRosenberg, E. (1998). Levels of analysis and the organization of affect. Review of General Psychology, 2<\/i>(3), 247\u2013270. doi:10.1037\/\/1089-2680.2.3.247 <\/span>\n\nRozin, P., & Cohen, A. (2003). High frequency of facial expressions corresponding to confusion, concentration, and worry in an analysis of maturally occurring facial expressions of Americans. Emotion, 3<\/i>, 68\u201375. <\/span>\n\nSabourin, J., Mott, B., & Lester, J. (2011). Modeling learner affect with theoretically grounded dynamic Bayesian networks. In S. D'Mello, A. Graesser, B. Schuller, & J. Martin (Eds.), Proceedings of the 4th International Conference on Affective Computing and Intelligent Interaction <\/i>(pp. 286\u2013295). Berlin\/Heidelberg: Springer-Verlag. <\/span>\n\nSan Pedro, M., Baker, R. S., Bowers, A. J., & Heffernan, N. T. (2013). Predicting college enrollment from student interaction with an intelligent tutoring system in middle school. In S. D'Mello, R. Calvo, & A. Olney (Eds.), Proceedings of the 6th International Conference on Educational Data Mining <\/i>(EDM 2013) (pp. 177\u2013184). International Educational Data Mining Society. <\/span>\n\nSchwarz, N. (2012). Feelings-as-information theory. In P. Van Lange, A. Kruglanski, & T. Higgins (Eds.), Handbook of theories of social psychology <\/i>(pp. 289\u2013308). Thousand Oaks, CA: Sage. <\/span>\n\nShute, V. J., Ventura, M., & Kim, Y. J. (2013). Assessment and learning of qualitative physics in Newton's playground. The Journal of Educational Research, 106<\/i>(6), 423\u2013430. <\/span>\n\nSinha, T., Jermann, P., Li, N., & Dillenbourg, P. (2014). Your click decides your fate: Inferring information processing and attrition behavior from MOOC video clickstream interactions. Paper presented at the Empirical Methods in Natural Language Processing: Workshop on Modeling Large Scale Social Interaction in Massively Open Online Courses. http:\/\/www.aclweb.org\/anthology\/W14-4102 <\/span>\n\nStrain, A., & D'Mello, S. (2014). Affect regulation during learning: The enhancing effect of cognitive reappraisal. Applied Cognitive Psychology, 29<\/i>(1), 1\u201319. doi:10.1002\/acp.3049<\/span>\n\nTracy, J.L. (2014). An evolutionary approach to understanding distinct emotions. Emotion Review, 6<\/i>(4), 308\u2013 312. <\/span>\n\nVanLehn, K., Siler, S., Murray, C., Yamauchi, T., & Baggett, W. (2003). Why do only some events cause learning during human tutoring? Cognition and Instruction, 21<\/i>(3), 209\u2013249. doi:10.1207\/S1532690XCI2103_01 <\/span>\n\nWang, Z., Miller, K., & Cortina, K. (2013). Using the LENA in teacher training: Promoting student involvement through automated feedback. Unterrichtswissenschaft, 4<\/i>, 290\u2013305. <\/span>\n\nWen, M., Yang, D., & Ros\u00e9, C. (2014). Sentiment analysis in MOOC discussion forums: What does it tell us? In J. Stamper, S. Pardos, M. Mavrikis, & B. M. McLaren (Eds.), Proceedings of the 7th International Conference on Educational Data Mining <\/i>(EDM 2014), 4\u20137 July, London, UK (pp. 130\u2013137): International Educational Data Mining Society. http:\/\/www.cs.cmu.edu\/~mwen\/papers\/edm2014-camera-ready.pdf <\/span>\n\nWoolf, B., Burleson, W., Arroyo, I., Dragon, T., Cooper, D., & Picard, R. (2009). Affect-aware tutors: Recognizing and responding to student affect. International Journal of Learning Technology, 4<\/i>(3\/4), 129\u2013163. <\/span>\n\nYang, D., Wen, M., Howley, I., Kraut, R., & Ros\u00e9, C. (2015). Exploring the effect of confusion in discussion forums of massive open online courses. Proceedings of the 2nd ACM conference on Learning@Scale <\/i>(L@S 2015), 14\u201318 March 2015, Vancouver, BC, Canada (pp. 121\u2013130). New York: ACM. doi:10.1145\/2724660.2724677 <\/span>\n\n<\/a> Zeng, Z., Pantic, M., Roisman, G., & Huang, T. (2009). A survey of affect recognition methods: Audio, visual, and spontaneous expressions. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31<\/i>(1), 39\u201358.<\/span>\n\n
\n\n
\n\n1<\/a> orj. detector<\/span>\n\n<\/div>\n
\n\n2<\/a> orj. killer app<\/span>\n\n<\/div>\n","rendered":"

<\/a> Sidney K. D Mello<\/span><\/span><\/p>\n

Psikoloji ve Bilgisayar Bilimleri ve M\u00fchendisli\u011fi B\u00f6l\u00fcmleri, Notre Dame \u00dcniversitesi, ABD<\/span><\/span><\/p>\n

DOI: 10.18608\/hla17.010<\/span><\/span><\/p>\n

\u00d6Z<\/h2>\n

Bu b\u00f6l\u00fcm, duygular\u0131n \u00f6\u011frenmenin yayg\u0131nl\u0131\u011f\u0131n\u0131 ve \u00f6nemini tart\u0131\u015fmaktad\u0131r. Ke\u015fif odakl\u0131, veri g\u00fcd\u00fcml\u00fc, \u00f6\u011frenme analiti\u011fi (\u00d6A) ve e\u011fitsel veri madencili\u011fi (EVM) ile duygusal ve \u00f6\u011frenme bilimlerindeki teorik geli\u015fmeler ve metodolojileri birle\u015ftirerek kayda de\u011fer ilerleme sa\u011flanabilece\u011fini savunuyor. Bu alanlar\u0131n kesi\u015fimindeki temel, ortaya \u00e7\u0131kan ve gelecekteki ara\u015ft\u0131rma temalar\u0131 tart\u0131\u015f\u0131lmaktad\u0131r.<\/span><\/p>\n

Anahtar kelimeler<\/span>: Etkiler, duyu\u015fsal bilim, duyu\u015fsal bilgi i\u015flem, e\u011fitsel veri madencili\u011fi<\/span><\/p>\n

Bir makalem i\u00e7in bir hakemin \u201c\u00f6nerisiyle\u201d (D’Mello, 2016), son d\u00f6nemde (benim i\u00e7in) yeni bir istatistiksel y\u00f6ntem olan genelle\u015ftirilmi\u015f toplan\u0131r karma modelleri \u00f6\u011frenmeye ba\u015flad\u0131m GTKM; McKeown ve Sneddon, 2014). GTKM bir cevap de\u011fi\u015fkenini (rezid\u00fcel) art\u0131klar aras\u0131ndaki oto d\u00fczeltmelere de\u011finerek (zaman serileri verisinde), \u00f6ng\u00f6r\u00fcsel de\u011fi\u015fkenlere ait parametrik ve parametrik olmayan d\u00fczg\u00fcn fonksiyonlar\u0131n toplan\u0131r bir birle\u015fimi ile modellemeyi ama\u00e7lar. \u0130lk ba\u015fta, bu makale i\u00e7in biraz daha fazla \u00e7al\u0131\u015fma d\u00fc\u015f\u00fcncesi beni biraz ho\u015fnutsuz k\u0131ld\u0131. Endi\u015fem son d\u00fczeltme tarihine kadar yeni bir metodu \u00f6\u011frenmek ve uygulamak i\u00e7in yeterli zaman\u0131m\u0131n olmayaca\u011f\u0131na dair d\u00fc\u015f\u00fcncemden kaynakland\u0131. Hi\u00e7bir \u015fey yapmad\u0131m. Son tarih yakla\u015f\u0131rken kayg\u0131 hafif pani\u011fe d\u00f6n\u00fc\u015ft\u00fc. Son olarak tavsiye edilen bir makaleyi indirerek GTKM’lere bakmaya karar verdim. Makale g\u00f6z al\u0131c\u0131 grafiklere sahipti, bu da merak duygumu uyand\u0131rd\u0131 ve beni daha fazla ke\u015ffetmeye motive etti. Merak, metot hakk\u0131nda daha \u00e7ok okuduk\u00e7a h\u0131zl\u0131 bir \u015fekilde ilgiye son olarak da yakla\u015f\u0131m\u0131n g\u00fcc\u00fcn\u00fc fark etti\u011fimde de heyecana d\u00f6n\u00fc\u015ft\u00fc. Bu bir \u015feyler anlam ifade etmedi\u011finde kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131<\/i><\/span> ve hayal k\u0131r\u0131kl\u0131\u011f\u0131,<\/i><\/span> neredeyse pes edecekken umutsuzluk, ilerleme kaydetti\u011fimi d\u00fc\u015f\u00fcnd\u00fc\u011f\u00fcmde umut ve son olarak da ger\u00e7ekten bir ilerleme kaydetti\u011fimde ise mutluluk ve haz gibi baz\u0131 yo\u011fun duygulara neden olarak beni teknik detaylarda bata \u00e7\u0131ka ilerlemeye motive etti. Daha sonra baz\u0131 R s\u00f6z dizimi \u00fczerinde de\u011fi\u015fiklikler yaparak metodu kendi verim \u00fczerinde uygulamaya koyuldum. Umut<\/i><\/span>, haz ve mutlulu\u011fun<\/i><\/span> aras\u0131na daha fazla hayal k\u0131r\u0131kl\u0131\u011f\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve <\/i><\/span>umutsuzluk serpi\u015ftirildi<\/i><\/span>. Sonunda hepsini \u00e7al\u0131\u015ft\u0131rd\u0131m ve sonu\u00e7lar\u0131 yazd\u0131m. Yazma ve d\u00fczeltme d\u00f6ng\u00fclerinde baz\u0131 ba\u015fka duygular daha olu\u015ftu. Sonunda ba\u015fard\u0131m. G\u00f6n\u00fcl rahatl\u0131\u011f\u0131, hafifleme<\/i><\/span> ve biraz da \u00f6v\u00fcn\u00e7<\/i><\/span> hissediyordum. Bu \u00f6rne\u011fin de g\u00f6sterdi\u011fi gibi, \u00f6\u011frenme s\u00fcreci boyunca bir duygular dip dalgas\u0131 vard\u0131r. Bu t\u00fcm \u201cbili\u015fin\u201d \u201cduygularla\u201d ili\u015fkili oldu\u011fu \u00f6\u011frenmeye has de\u011fildir. Duygular her zaman bilin\u00e7li olarak deneyimlenmeyebilirler (Ohman ve Soares, 1994) ancak yine de vard\u0131rlar ve bili\u015fi etkilerler. Ayn\u0131 zamanda, duygular bir hava bo\u015flu\u011funda olu\u015fmazlar, \u00f6\u011frenmenin sosyal kuma\u015f\u0131yla derinden sarmalanm\u0131\u015f vaziyettedirler. En temel i\u015fi \u00f6\u011frenmek olan tipik bir \u00f6\u011frenen taraf\u0131ndan hangi duygular yelpazesinin deneyimlenece\u011fini hayal etmek \u00e7ok zor de\u011fildir. Pekrun ve Stephens (2011), bunlara \u201cakademik duygular\u201d demi\u015f ve onlar\u0131 d\u00f6rt grupta s\u0131n\u0131fland\u0131rm\u0131\u015ft\u0131r. Ba\u015far\u0131 duygular\u0131 (g\u00f6n\u00fcl rahatl\u0131\u011f\u0131, endi\u015fe ve hayal k\u0131r\u0131kl\u0131\u011f\u0131), \u00f6\u011frenme etkinlikleri (\u00f6dev, bir teste girme) ve \u00e7\u0131kt\u0131larla (ba\u015far\u0131, ba\u015far\u0131s\u0131zl\u0131k) ba\u011flant\u0131l\u0131d\u0131r. Konu ba\u015fl\u0131\u011f\u0131<\/i><\/span> duygular\u0131 \u00f6\u011frenme i\u00e7eri\u011fi (klasik edebiyat okurken hik\u00e2yenin kahraman\u0131 ile duyguda\u015fl\u0131k kurmak) ile uyumludur. Sosyal duygular \u00f6v\u00fcn\u00e7, utan\u00e7 ve k\u0131skan\u00e7l\u0131k da vard\u0131r \u00e7\u00fcnk\u00fc e\u011fitim sosyal ortamlarda yer al\u0131r. Son olarak, \u00f6zg\u00fcnl\u00fckle kar\u015f\u0131la\u015f\u0131ld\u0131\u011f\u0131nda \u015fa\u015fk\u0131nl\u0131k veya bir a\u00e7mazla kar\u015f\u0131la\u015f\u0131ld\u0131\u011f\u0131nda kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 olmas\u0131 gibi epistemik<\/i><\/span> duygular bili\u015fsel s\u00fcre\u00e7lerden do\u011farlar.<\/p>\n

Duygular sadece tesad\u00fcfi de\u011fildirler veya evrimle\u015fmemi\u015ftirler (Darwin, 1872; Tracy, 2014). Duygular bilgi ile ilgili problemleri (kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131), uyar\u0131lma ile ilgili problemleri (b\u0131kk\u0131nl\u0131k), yakla\u015fan performansla ilgili meseleleri (endi\u015fe) ve kolayl\u0131kla a\u015f\u0131lamayacak zorluklar\u0131 (hayal k\u0131r\u0131kl\u0131\u011f\u0131) vurgulayarak uyar\u0131 verme i\u015flevini<\/i><\/span> g\u00f6r\u00fcrler (Schwarz, 2012). \u0130nsanlar\u0131n bir olaya de\u011feri hedef uygunlu\u011fu ve hedef e\u015fle\u015fmesi a\u00e7\u0131s\u0131ndan de\u011fer bi\u00e7ti\u011fi bir para birimi g\u00f6revi g\u00f6ren de\u011ferlendirici i\u015flevler<\/i><\/span> icra ederler. (Izard, 2010). Duygular bili\u015fsel oda\u011f\u0131 s\u0131n\u0131rlayarak ya da geni\u015fleterek olumlu duygularla daha geni\u015f, yukar\u0131dan a\u015fa\u011f\u0131ya ve \u00fcretici i\u015flemeyi (geni\u015fletilmi\u015f odak) desteklemeye (Barth ve Funke, 2010; Schwarz, 2012) k\u0131yasla, olumsuz duygularla i\u015flemenin dar, a\u015fa\u011f\u0131dan yukar\u0131ya ve odaklanm\u0131\u015f modellerini<\/i><\/span> (s\u0131n\u0131rl\u0131 odak) olu\u015fturarak ge\u00e7i\u015f i\u015flevini icra ederler (Fredrickson ve Branigan, 2005; Isen, 2008). Ger\u00e7ekten de duygular bellek, problem \u00e7\u00f6zme, karar lama ve bili\u015fin di\u011fer alanlar\u0131 \u00fczerindeki etkilerinde a\u00e7\u0131k\u00e7a g\u00f6r\u00fcld\u00fc\u011f\u00fc gibi d\u00fc\u015f\u00fcnceye h\u00e2kimdirler (detayl\u0131 inceleme i\u00e7in, bk. Clore ve Huntsinger, 2007).<\/p>\n

Peki \u201cduygu\u201d tam olarak nedir? Do\u011frusunu s\u00f6ylemek gerekirse, ger\u00e7ekten bilmiyoruz veya en az\u0131ndan tam olarak uzla\u015fam\u0131yoruz (Izard, 2010). Bu durum duygunun psikolojik temellerine dair en g\u00fcncel tart\u0131\u015fmalardan da -bazen “100 ya\u015f\u0131ndaki duygu sava\u015f\u0131” diye de adland\u0131r\u0131lan- rahatl\u0131kla anla\u015f\u0131labilir (Lench, Bench ve Flores, 2013; Lindquist, Siegel, Quigley ve Barrett, 2013). Neyse ki, baz\u0131 konularda genel bir anla\u015fma sa\u011flanm\u0131\u015ft\u0131r. Duygular beyin- beden- \u00e7evre etkile\u015fiminden olu\u015fan kavramsal varl\u0131klard\u0131r. Fakat onlar\u0131 beyin, beden veya \u00e7evreye bakarak bulamazs\u0131n\u0131z. Tam tersine, organizma ve \u00e7evre etkile\u015fimleri, \u00e7oklu zaman \u00f6l\u00e7ekleri ve n\u00f6robiyolojik, fizyolojik ve davran\u0131\u015fsal olarak d\u0131\u015favurumcu, eylem odakl\u0131 ve bili\u015fsel\/ bili\u015f\u00fcst\u00fc\/\u00f6znel gibi \u00e7oklu d\u00fczeyler aras\u0131ndaki de\u011fi\u015fimleri tetikledi\u011finde duygular ortaya \u00e7\u0131kar (Lewis, 2005). \u201cDuygu\u201d s\u00fcregelen durumsal ba\u011flam taraf\u0131ndan ayarlanarak bu de\u011fi\u015fimlere yans\u0131t\u0131l\u0131r. Ayn\u0131 duygusal kategori (\u00f6r. endi\u015fe) tetikleyici olaya (Tracy, 2014), biyolojik\/bili\u015fsel\/bili\u015f\u00fcst\u00fc s\u00fcre\u00e7lere (Gross, 2008; Moors, 2014) ve sosyok\u00fclt\u00fcrel etkilere ba\u011fl\u0131 olarak (Mesquita ve Boiger, 2014; Parkinson, Fischer ve Manstead, 2004) farkl\u0131 \u015fekillerde a\u00e7\u0131\u011fa \u00e7\u0131kar. \u00d6rne\u011fin, belirli ko\u015fullara ba\u011fl\u0131 olarak endi\u015feye sebep olan bir olay (topluluk \u00f6n\u00fcnde konu\u015fma, s\u0131nava girme), zamansal ba\u011flam (konu\u015fmadan bir g\u00fcn ya da bir dakika \u00f6nce), n\u00f6robiyolojik sistem (s\u0131n\u0131r \u00e7izgi uyar\u0131lmas\u0131) ve sosyal ba\u011flam (i\u015f arkada\u015flar\u0131n\u0131n ya da bir yabanc\u0131n\u0131n \u00f6n\u00fcnde konu\u015fmak) endi\u015fenin farkl\u0131 \u201ck\u0131s\u0131mlar\u0131n\u0131 tetikleyecektir. Farkl\u0131l\u0131klar\u0131n ve de\u011fi\u015fkenli\u011fin bu d\u00fczeyi insanlar ve duygular\u0131 dinamik ve uyarlanabilir oldu\u011fundan beklendiktir. De\u011fi\u015fmez duygular\u0131n evrimsel de\u011feri \u00e7ok azd\u0131r.<\/p>\n

\u00d6\u011frenme analitikleri (\u00d6A) ve e\u011fitsel veri madencili\u011fi (EVM) nereye uygundur? Bir taraftan, duygular\u0131n \u00f6\u011frenmedeki merkezi rol\u00fc g\u00f6z \u00f6n\u00fcnde tutuldu\u011funda, \u00f6\u011frenmeyi duygular\u0131 dikkate almayarak analiz etme giri\u015fimleri tamamlanmam\u0131\u015f olacakt\u0131r. Di\u011fer taraftan, duygusal fenomenlerin karma\u015f\u0131kl\u0131\u011f\u0131 ve belirsizli\u011fi d\u00fc\u015f\u00fcn\u00fcld\u00fc\u011f\u00fcnde, \u00f6\u011frenme s\u0131ras\u0131nda olu\u015fan duygular\u0131 \u00d6A ve EVM y\u00f6ntemleri olmadan analiz etme giri\u015fimleri sadece s\u0131\u011f i\u00e7g\u00f6r\u00fcler getirecektir. Neyse ki, \u00f6\u011frenme \u00fcr\u00fcnleri s\u00fcre\u00e7lerindeki duygular\u0131n olu\u015f s\u0131kl\u0131\u011f\u0131 ve etkilerini \u00e7al\u0131\u015fmak i\u00e7in veriye dayal\u0131 analitik yakla\u015f\u0131m\u0131 benimseyen bir \u00e7al\u0131\u015fma alan\u0131 vard\u0131r. Bu b\u00f6l\u00fcmde ben bu disiplinler aras\u0131 ara\u015ft\u0131rma alan\u0131ndaki temel, geli\u015fmekte olan ve gelecek temalar\u0131n baz\u0131lar\u0131n\u0131 vurguluyorum.<\/p>\n

Terminolojiye bir notla ba\u015flayal\u0131m. Duygular motivasyon, tutumlar, tercihler, fizyoloji, uyar\u0131lma ve ona at\u0131fta bulunulmakta kullan\u0131lan di\u011fer yap\u0131lar k\u00fcmesiyle ili\u015fiklidir ancak e\u015f de\u011fer de\u011fildir. Duygular ayn\u0131 zamanda miza\u00e7 ve duyu\u015fsal \u00f6zelliklerden de ayr\u0131d\u0131r (Rosenberg, 1998). Duygular hislerle de ayn\u0131 de\u011fildir. A\u00e7l\u0131k bir histir ancak duygu de\u011fildir. Ac\u0131 da de\u011fil. Duygunun ne oldu\u011funa dair bir ihtilaf da vard\u0131r. \u00d6fke kesinlikle bir duygudur ancak ya \u015fa\u015fk\u0131nl\u0131k? \u015ea\u015f\u0131rman\u0131n duyu\u015fsal bile\u015fenleri vard\u0131r (\u015fa\u015f\u0131rmaya dair hisler, y\u00fcz ifadesi karakteristikleri; (D’Mello ve Graesser, 2014b), fakat onun bir duygu olup olmad\u0131\u011f\u0131na dair bir miktar tart\u0131\u015fma bulunmaktad\u0131r (Hess, 2003; Rozin ve Cohen, 2003). Dolay\u0131s\u0131yla bu b\u00f6l\u00fcm\u00fcn geri kalan\u0131nda, daha k\u0131s\u0131tlay\u0131c\u0131 bir terim olan \u201cduygu\u201d dan ziyade daha kapsay\u0131c\u0131 bir terim olan \u201cduyu\u015fsal durumu\u201d kullanaca\u011f\u0131m.<\/p>\n

ANA TEMALAR<\/h2>\n

\u00d6\u011frenmedeki duyu\u015fu \u00e7al\u0131\u015fma amac\u0131yla \u00d6A\/EVM metotlar\u0131n\u0131 kullan\u0131m\u0131n\u0131 vurgulamak i\u00e7in a\u015fa\u011f\u0131daki d\u00f6rt temay\u0131 se\u00e7tim. \u00dcst\u00fcnk\u00f6r\u00fc bir bi\u00e7imde bir\u00e7ok \u00e7al\u0131\u015fmay\u0131 g\u00f6zden ge\u00e7irmektense her temada bir ya da iki \u00f6rnek te\u015fkil eden \u00e7al\u0131\u015fmay\u0131 bir belirli bir d\u00fczeyde inceliyorum. Bu da bir\u00e7ok harika \u00e7al\u0131\u015fmadan bahsedilmeyece\u011fi anlam\u0131n geliyor, fakat ben bu her tema i\u00e7in \u00e7al\u0131\u015fma alan\u0131n\u0131 ara\u015ft\u0131rma i\u015fini okuyucuya b\u0131rak\u0131yorum. Ben s\u00fcreci desteklemek ad\u0131na uygun oldu\u011funda, derleme \u00e7al\u0131\u015fmalar\u0131 \u00f6nerece\u011fim.<\/p>\n

T\u0131klama Ak\u0131\u015f\u0131 Verisinden Duyu\u015fsal Analiz<\/h3>\n

\u00d6A\/EVM tekniklerinin en temel kullan\u0131mlar\u0131ndan bir tanesi, \u00f6\u011frenenlerin bili\u015fsel s\u00fcre\u00e7lerini anlamak i\u00e7in \u00f6\u011frenme teknolojileri ile etkile\u015fimlerden olu\u015fturulan zengin veri ak\u0131\u015f\u0131n\u0131 kullanmakt\u0131r (Corbett ve Anderson, 1995; Sinha, Jermann, Li ve Dillenbourg, 2014). A\u015fa\u011f\u0131daki \u00e7al\u0131\u015fmada da belirtildi\u011fi \u00fczere, duyu\u015f kar\u0131\u015f\u0131ma eklendi\u011finde tamamlay\u0131c\u0131 bir dizi i\u00e7g\u00f6r\u00fc elde edilebilir.<\/p>\n

Bosch ve D’Mello (bas\u0131m a\u015famas\u0131nda) \u00f6\u011frencilerin ilk programlama oturumlar\u0131 s\u00fcresince duyu\u015fsal deneyimleri \u00fczerine bir laboratuar \u00e7al\u0131\u015fmas\u0131 yapt\u0131lar. \u00c7aylak \u00f6\u011frencilerden (N=99) 25 dakikal\u0131k desteklenmi\u015f bir \u00f6\u011frenme evresi ve 10 dakikal\u0131k desteksiz bir kaybolma evresini i\u00e7eren \u00f6z y\u00f6netimli bilgisayarl\u0131 \u00f6\u011frenme ortam\u0131n\u0131 kullanarak, Python dilinde bilgisayar programlaman\u0131n temellerini \u00f6\u011frenmeleri istendi. T\u00fcm \u00f6\u011fretimsel etkinlikler (kodlama, metin okuma, kodu test etme, hatalar\u0131 alma vb.) sistem g\u00fcnl\u00fc\u011f\u00fcne kaydedildi ve \u00f6\u011frencilerin y\u00fczlerinin videolar\u0131 ve bilgisayar ekranlar\u0131 kaydedildi. \u00d6\u011frenciler \u00f6\u011frenme oturumunun hem en ard\u0131ndan bu videolar\u0131 izleme esnas\u0131nda yakla\u015f\u0131k 100 puanda (her 15 saniyede bir) ge\u00e7mi\u015fe dair duyu\u015fsal yarg\u0131 protokol\u00fc arac\u0131l\u0131\u011f\u0131yla duyu\u015fsal h\u00fck\u00fcmler verdiler (Porayska-Pomsta, Mavrikis, D’Mello, Conati ve Baker, 2013). \u0130lgilenilen duyu\u015fsal durumlar \u00f6fke, endi\u015fe, can s\u0131k\u0131nt\u0131s\u0131, merak, i\u011frenme, korku, hayal k\u0131r\u0131kl\u0131\u011f\u0131, ak\u0131\u015f\/ me\u015fguliyet, mutluluk, \u00fcz\u00fcnt\u00fc ve \u015fa\u015fk\u0131nl\u0131k idi. Sadece me\u015fguliyet, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131, can s\u0131k\u0131nt\u0131s\u0131 ve merak yeterli s\u0131kl\u0131kta olu\u015farak daha ileri analizi gerek\u00e7elendirebildi.<\/p>\n

Yazarlar etkile\u015fim olaylar\u0131n\u0131n duygusal durumlara nas\u0131l yol a\u00e7t\u0131\u011f\u0131n\u0131 ve duygusal durumlar\u0131n \u00e7e\u015fitli davran\u0131\u015flar\u0131 nas\u0131l tetikledi\u011fini incelemi\u015ftir. Yazarlar etkile\u015fim olaylar\u0131n\u0131n duyu\u015fsal durumlara nas\u0131l sebebiyet verdi\u011fini ve duyu\u015fsal durumlar\u0131n farkl\u0131 davran\u0131\u015flar\u0131 nas\u0131l tetikledi\u011fini ara\u015ft\u0131rd\u0131lar Her bir \u00f6\u011frenen i\u00e7in desteklenmi\u015f \u00f6\u011frenme evresi boyunca etkile\u015fim olaylar\u0131 (t\u0131klama ak\u0131\u015f verisi) ve duyu\u015fsal durumlar\u0131 (\u00f6z raporlar) aras\u0131na serpi\u015ftiren zaman serileri olu\u015fturdular. Zaman serisi modelleme teknikleri (D’Mello, Taylor ve Graesser, 2007), duygusal durumlar ile etkile\u015fim olaylar\u0131 aras\u0131ndaki \u00f6nemli ge\u00e7i\u015fleri tan\u0131mlamak i\u00e7in kullan\u0131lm\u0131\u015ft\u0131r.<\/p>\n

\"\"<\/p>\n

<\/a> \u015eekil 10.1. Duygusal durumlar ile bilgisayar programlama bili\u015fsel destek \u00f6\u011frenme s\u0131ras\u0131nda etkile\u015fim olaylar\u0131 aras\u0131ndaki \u00f6nemli ge\u00e7i\u015fler. D\u00fcz \u00e7izgiler duygu ge\u00e7i\u015fleri de d\u00e2hil ge\u00e7i\u015fleri g\u00f6stermektedir. Kesik \u00e7izgiler duygusal durumlar\u0131 i\u00e7ermeyen ge\u00e7i\u015fler g\u00f6sterir. Problemi G\u00f6ster: yeni bir al\u0131\u015ft\u0131rmaya ba\u015flama; Okuma: y\u00f6nergeleri ve\/ veya problem durumunu g\u00f6rme; Kodlama: mevcut kodu d\u00fczenleme veya g\u00f6rme; \u0130pucunuG\u00f6ster: ipucunu g\u00f6rme; TestRunError: kod \u00e7al\u0131\u015ft\u0131r\u0131lm\u0131\u015f ve s\u00f6z dizimi ya da \u00e7al\u0131\u015fma zaman\u0131 hatas\u0131 ile kar\u015f\u0131la\u015f\u0131lm\u0131\u015f; TestRunSuccess: kod s\u00f6z dizimi hatas\u0131 ya da \u00e7al\u0131\u015fma zaman\u0131 hatas\u0131 olmadan \u00e7al\u0131\u015ft\u0131r\u0131lm\u0131\u015f (ancak do\u011fruluk i\u00e7in kontrol yap\u0131lmam\u0131\u015f); G\u00f6nderiHatas\u0131: kod g\u00f6nderildi ve bir hata veya yanl\u0131\u015f bir cevap \u00fcretti; G\u00f6nderiBa\u015far\u0131l\u0131: kod g\u00f6nderildi ve do\u011fruydu.<\/i><\/span><\/p>\n

Sonu\u00e7ta olu\u015fan y\u00f6nl\u00fc \u00e7izge \u015eekil 10.1 de verilmi\u015ftir. Duyu\u015fsal durumlar\u0131 i\u00e7ermeyen etkile\u015fim olaylar\u0131 aras\u0131nda baz\u0131 ge\u00e7i\u015fler bulunmaktayd\u0131 (kesikli \u00e7izgiler). Bu di\u011fer etkile\u015fim olaylar\u0131na nazaran (saniyede 1 s\u0131kl\u0131kta) duygu \u00f6rneklem al\u0131m\u0131n\u0131n az s\u0131kl\u0131kta olu\u015funa (her 15 saniyede bir) ba\u011fl\u0131yd\u0131.<\/p>\n

Daha ilgi \u00e7ekici ge\u00e7i\u015fler duyu\u015fsal durumlar\u0131 i\u00e7ermekteydi. \u00d6zellikle, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131ndan \u00f6nce bir yanl\u0131\u015f bir cevap g\u00f6nderme (G\u00f6nderiHatas\u0131) gelmekte; daha sonra bu duyu\u015fsal durumlar\u0131 bir ipucu talebi (\u0130pucunuG\u00f6ster) veya kodu yap\u0131land\u0131rma (Kodlama) talebi izlemekteydi ki bu durumlar\u0131n kendisi de bir kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131 tetiklemekteydi. \u00d6\u011fretim metinlerinin okunmas\u0131 (problem tan\u0131mlamalar\u0131 d\u00e2hil) kat\u0131l\u0131m, merak, can s\u0131k\u0131nt\u0131s\u0131 ve kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n da belirtisiydi ancak hayal k\u0131r\u0131kl\u0131\u011f\u0131na u\u011framad\u0131. Bir ba\u015fka deyi\u015fle, t\u00fcm anahtar duyu\u015fsal durumlar bilgi \u00f6z\u00fcmseme (okuma) ve yap\u0131land\u0131rma (kodlama) etkinlikleriyle ili\u015fkiliydi. Ancak sadece b\u00fcy\u00fck ihtimalle \u00f6\u011frenme f\u0131rsatlar\u0131 olan kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131 ba\u015far\u0131s\u0131zl\u0131\u011fa (G\u00f6nderiHatas\u0131) ve sonras\u0131nda gelen yard\u0131m isteme davran\u0131\u015flar\u0131na (\u0130pucunu G\u00f6ster) e\u015flik etti. Birlikte bak\u0131ld\u0131\u011f\u0131nda, ge\u00e7i\u015f modeli g\u00fc\u00e7 durumlar ve sonu\u00e7ta olu\u015fan kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 veya hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131 etkinle\u015ftirici olumsuz durumlar\u0131n \u00f6\u011frenmedeki \u00f6nemli rol\u00fcn\u00fc vurgulamaktad\u0131r. (D’Mello ve Graesser, 2012b; VanLehn, Siler, Murray, Yamauchi ve Baggett, 2003). Ayn\u0131 zamanda, duygular\u0131n \u00f6\u011frenme s\u00fcrecinin i\u00e7erisine nas\u0131l serpi\u015ftirilmi\u015f oldu\u011funu a\u00e7\u0131klamaktad\u0131r.<\/p>\n

Etkile\u015fim \u015eemalar\u0131ndan Duygu Saptama<\/h3>\n

Duyu\u015fsal durumlar kavramsal varl\u0131klar (yap\u0131lar) oldu\u011fundan do\u011frudan do\u011fruya \u00f6l\u00e7\u00fclemezler. Ancak \u00e7evre-ki\u015fi etkile\u015fimlerinden do\u011farlar ve bili\u015fin de\u011fi\u015ftirilmesiyle eylemi etkilerler. Bu nedenle, g\u00f6zler \u00f6n\u00fcne serilen ba\u011flam ve \u00f6\u011frenen eylemleri analiz edilerek duyguyu “\u00e7\u0131karsamak” m\u00fcmk\u00fcn olabilmelidir. Etkile\u015fim temelli”, “kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyas\u0131 tabanl\u0131” veya “alg\u0131lay\u0131c\u0131s\u0131z” olarak adland\u0131r\u0131lan duygu saptama \u00e7al\u0131\u015fma alan\u0131 on y\u0131ldan uzun bir s\u00fcre \u00f6nce ba\u015flam\u0131\u015ft\u0131r. (Ai vd., 2006; D’Mello, Craig, Sullins ve Graesser, 2006) ve yak\u0131n zamanda Baker ve Ocumpaugh (2015) taraf\u0131ndan derlenmi\u015ftir.<\/p>\n

\u00d6rnek olarak, ortaokul -ve lise- matemati\u011fi i\u00e7in bir Ak\u0131ll\u0131 \u00d6\u011fretim Sistem (A\u00d6S) olan ve ABD’de ola\u011fan matematik \u00f6\u011fretiminin bir par\u00e7as\u0131 olarak yakla\u015f\u0131k 50, 0000 \u00f6\u011frenci taraf\u0131ndan kullan\u0131lan, ASSISTments’a duygu saptay\u0131c\u0131lar geli\u015ftiren Pardos, Baker, San Pedro ve Gowda (2013) d\u00fc\u015f\u00fcn\u00fclebilir. (Razzaq vd., 2005). Yazarlar otomatikle\u015ftirilmi\u015f duygu saptay\u0131c\u0131lar1<\/sup><\/a> olu\u015fturmak i\u00e7in denetimli bir \u00f6\u011frenme yakla\u015f\u0131m\u0131 benimsediler. Okul bilgisayar laboratuvarlar\u0131nda ASSISTments kullanan 229 \u00f6\u011frenciden e\u011fitim verisi toplad\u0131lar. \u0130nsan g\u00f6zlemciler Baker-Rodrigo G\u00f6zlem Metodu Protokol\u00fcn\u00fc kullanarak (BRGYP) \u00f6\u011frencilerin ASSISTments ile etkile\u015fimlerindeki \u00e7evrimi\u00e7i duygu g\u00f6zlemlerini (k\u0131sa ek a\u00e7\u0131klamalar) getirdiler (Ocumpaugh, Baker ve Rodrigo, 2012). Bu protokole g\u00f6re e\u011fitim alm\u0131\u015f g\u00f6zlemciler, g\u00f6zlemlenebilir davran\u0131\u015f temel al\u0131narak aray\u00fcze y\u00f6nelik a\u00e7\u0131k eylemler, akranlar ve \u00f6\u011fretmenle etkile\u015fim, beden, hareketler, el hareketleri ve y\u00fcz ifadeleri de d\u00e2hil olmak \u00fczere duyguya dair ger\u00e7ek zamanl\u0131 ek a\u00e7\u0131klamalar sa\u011flarlar. G\u00f6zlemciler d\u00f6rt duyu\u015fsal durum (can s\u0131k\u0131nt\u0131s\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131 ba\u011flanm\u0131\u015f konsantrasyon ve kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131) ve iki davran\u0131\u015f (etkinlikle ilgisiz \u015feyler yapmak veya sistemi oyunla\u015ft\u0131rmak) kodlad\u0131lar. ASSISTments kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyalar\u0131ndan \u00e7\u0131kart\u0131lan baz\u0131 \u00f6zellikler kullan\u0131larak her bir duyu\u015fsal durumu di\u011fer durumlardan ay\u0131racak (\u00f6r. can s\u0131k\u0131nt\u0131s\u0131na kar\u015f\u0131 di\u011ferleri) denetimli \u00f6\u011frenme teknikleri kullan\u0131ld\u0131. Etki tespiti do\u011frulu\u011fu, etki i\u00e7in “.632” ile “.678” [A-\u00fcss\u00fc metrik olarak \u00f6l\u00e7\u00fclm\u00fc\u015ft\u00fcr (EAA veya A\u0130KAA – al\u0131c\u0131 i\u015flem karakteristik e\u011frisi alt\u0131ndaki alana benzer)]- ve davran\u0131\u015flar i\u00e7in “.802” ila “.819” aras\u0131nda de\u011fi\u015fmi\u015ftir. S\u0131n\u0131fland\u0131r\u0131c\u0131n\u0131n do\u011frulanmas\u0131 e\u011fitim ve s\u0131nav verisi aras\u0131nda tam bir ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131 mecbur k\u0131larak, ayn\u0131 pop\u00fclasyondan yeni \u00f6\u011frencilere genellenebilirli\u011fi sa\u011flama \u015feklinde yap\u0131ld\u0131.<\/p>\n

Pardos vd. (2013) ayn\u0131 zamanda saptay\u0131c\u0131lar\u0131n\u0131n \u00f6ng\u00f6r\u00fcsel ge\u00e7erlili\u011fine dair ba\u015flang\u0131\u00e7 niteli\u011finde kan\u0131tlar sundular. Bu \u00f6l\u00e7e\u011fin geli\u015ftirilmesinden bir ka\u00e7 y\u0131l \u00f6nce 2004-2006 y\u0131llar\u0131 aras\u0131nda ASSISTments ile etkile\u015fimi olan 1393 adet farkl\u0131 bir grup \u00f6\u011frencini dosyalar\u0131na alg\u0131lay\u0131c\u0131lar\u0131n uygulanmas\u0131 ile yap\u0131lm\u0131\u015ft\u0131r. Standart hale gelmi\u015f s\u0131nav puanlar\u0131 ile otomatik olarak \u00f6l\u00e7\u00fclen duygu ve davran\u0131\u015flar k\u0131smen ili\u015fkili bulunmu\u015ftur.<\/p>\n

Sonras\u0131nda, San Pedro, Baker, Bowers ve Heffernan (2013) otomatik saptay\u0131c\u0131lara dayanarak \u00fcniversite kay\u0131tlar\u0131n\u0131 tahmin etme giri\u015fiminde bulundular. 2004’ten 2009’a kadar ASSISTments ile etkile\u015fime giren 3707 \u00f6\u011frencini mevcut kay\u0131t g\u00fcnl\u00fc\u011f\u00fc dosyalar\u0131na saptay\u0131c\u0131lar\u0131 uygulad\u0131lar. Bu \u00f6\u011frencilerin \u00fcniversite kay\u0131t bilgileri Ulusal \u00d6\u011frenci B\u00fcrosundan elde edildi. Y\u0131llar sonra, olduk\u00e7a etkileyici bir bulgu olarak otomatik olarak \u00f6l\u00e7\u00fclen duyu\u015fsal durumlar\u0131n \u00fcniversiteye kay\u0131t yapt\u0131rman\u0131n en manidar \u00f6ng\u00f6r\u00fcc\u00fclerinden biri oldu\u011fu ortaya konmu\u015ftur.<\/p>\n

Bedensel \u0130\u015faretlerden Duygu Saptama<\/h3>\n

Duygu eylem i\u00e7in bedensel cevap sistemlerini harekete ge\u00e7irdi\u011fi i\u00e7in \u015fekillendirici bir fenomendir. Bu \u00f6\u011frenen duygusunun makine taraf\u0131ndan okunabilir bedensel sinyallere dayanarak \u00e7\u0131karsanaca\u011f\u0131n\u0131 m\u00fcmk\u00fcn k\u0131lmal\u0131d\u0131r. Bir\u00e7ok derlemde de tart\u0131\u015f\u0131ld\u0131\u011f\u0131 gibi, bedensel i\u015fretlerin duygular\u0131 saptamada kullan\u0131m\u0131 \u00fczerine zengin bir \u00e7al\u0131\u015fma alan\u0131 bulunmaktad\u0131r (Calvo ve D’Mello, 2010; D’Mello ve Kory, 2015; Zeng, Pantic, Roisman ve Huang, 2009). Ara\u015ft\u0131rma ge\u00e7mi\u015fte kontroll\u00fc ortamlardaki etkile\u015fimlere odaklanmas\u0131na kar\u015f\u0131n, ara\u015ft\u0131rmac\u0131lar bu \u00e7al\u0131\u015fmay\u0131 bilgisayar destekli s\u0131n\u0131flar ba\u015fta olmak \u00fczere ger\u00e7ek d\u00fcnyaya ta\u015f\u0131maya ba\u015flad\u0131lar. A\u015fa\u011f\u0131daki g\u00f6zden ge\u00e7irilen \u00e7al\u0131\u015fma bizim ara\u015ft\u0131rma grubumuz ve i\u015fbirlikli \u00e7al\u0131\u015ft\u0131klar\u0131m\u0131z taraf\u0131ndan ortaya konan benzer bir eme\u011fi yans\u0131tmakta ancak okuyucu Arroyo vd. (2009) bilgisayar destekli s\u0131n\u0131flarda duygu saptamaya y\u00f6nelik \u00f6nc\u00fc niteli\u011findeki \u00e7al\u0131\u015fmaya y\u00f6nlendirilmi\u015ftir.<\/p>\n

Bosch, D’Mello, Baker, Ocumpaugh ve Shute (2016) bilgisayar destekli bir s\u0131n\u0131f\u0131n karma\u015f\u0131k ger\u00e7ek d\u00fcnyas\u0131nda y\u00fcze dair \u00f6zelliklerden duygunun otomatik saptanmas\u0131n\u0131 \u00e7al\u0131\u015fm\u0131\u015flard\u0131r. Bu \u00e7al\u0131\u015fmada, 137 orta\u00f6\u011fretim ve lise \u00f6\u011frencisi k\u00fc\u00e7\u00fck gruplar halinde her zamanki fizik \/ fizik bilimleri derslerinin bir par\u00e7as\u0131 olarak iki g\u00fcn boyunca 1.5-2 saat Fizik Oyun Bah\u00e7esi (Shute, Ventura ve Kim, 2013) adl\u0131 kavramsal bir fizik e\u011fitim oyunu oynad\u0131. E\u011fitimli g\u00f6zlemciler yukar\u0131da anlat\u0131lan ASSISSTments \u00e7al\u0131\u015fmas\u0131nda oldu\u011fu gibi BRGYP saha g\u00f6zlem protokol\u00fcn\u00fc kullanarak can s\u0131k\u0131nt\u0131s\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131, hayal k\u0131r\u0131kl\u0131\u011f\u0131, ba\u011flant\u0131l\u0131- konsantrasyon ve hazza dair canl\u0131 k\u0131sa ek a\u00e7\u0131klamalar\u0131 i\u015flediler (Pardos vd., 2013). G\u00f6zlemciler ayn\u0131 zamanda \u00f6\u011frenciler konuyla alakas\u0131z olduklar\u0131nda da not ald\u0131lar.<\/p>\n

\"\"<\/p>\n

<\/a> \u015eekil 10.2. Bilgisayar \u0130fade Tan\u0131ma Alet \u00e7antas\u0131 kullan\u0131larak y\u00fcz niteliklerinin otomatik taranmas\u0131. Sa\u011fdaki grafikler farkl\u0131 y\u00fcz niteliklerinin (\u00f6r. indirilmi\u015f ka\u015flar, gerilmi\u015f g\u00f6z kapaklar\u0131) muhtemel etkinle\u015ftirilmesini g\u00f6stermektedir.<\/i><\/span><\/p>\n

Oyun s\u0131ras\u0131nda \u00f6\u011frencilerin y\u00fczlerinin ve v\u00fccutlar\u0131n\u0131n \u00fcst k\u0131sm\u0131 videolar\u0131 kaydedildi ve etki a\u00e7\u0131klamalar\u0131yla senkronize edildi. Videolar, 19 y\u00fcz eylem biriminin (Ekman ve Friesen, 1978) (\u00f6r. y\u00fckseltilmi\u015f ka\u015f, gerilmi\u015f dudaklar), ba\u015f pozunun (uyumu) ve ba\u015f pozisyonunun olas\u0131l\u0131\u011f\u0131n\u0131n tahminini sa\u011flayan FACET bilgisayarl\u0131 g\u00f6r\u00fc\u015f program\u0131 (Emotient, 2014) kullan\u0131larak i\u015flendi (ekran g\u00f6r\u00fcnt\u00fcs\u00fc i\u00e7in bk. \u015eekil 10.2). Beden hareketleri hareket filtreleme algoritmalar\u0131 kullanarak tahmin edildi (Kory, D’Mello ve Olney, 2015) (bk. Fig\u00fcr 10.3). Hem y\u00fcz ifadeleri hem de bedensel hareketler kullan\u0131larak her bir duyu\u015fsal durumun saptay\u0131c\u0131lar\u0131 denetimli \u00f6\u011frenme y\u00f6ntemleri ile geli\u015ftirildi (\u00f6r. can s\u0131k\u0131nt\u0131s\u0131na kar\u015f\u0131 di\u011fer durumlar). Alg\u0131lay\u0131c\u0131lar duygu i\u00e7in .610’dan .867 aral\u0131\u011f\u0131 aras\u0131nda etkinlik d\u0131\u015f\u0131 davran\u0131\u015flar i\u00e7in ise .816 do\u011fruluklarla (miktarlar yukar\u0131da belirtilen EAA metrikleriyle \u00f6l\u00e7\u00fclerek) k\u0131smen ba\u015far\u0131l\u0131 oldular. Takip eden analizler duygu alg\u0131lay\u0131c\u0131lar\u0131n\u0131n \u00f6\u011frenciler, farkl\u0131 g\u00fcnler ve farkl\u0131 cinsiyet ve etnisiteler (insanlar taraf\u0131ndan alg\u0131land\u0131\u011f\u0131 \u015fekliyle) boyunca genelleme yapt\u0131\u011f\u0131n\u0131 do\u011frulad\u0131.<\/p>\n

\"\"<\/p>\n

<\/a> \u015eekil 10.3. Beden hareketlerinin hareket sil\u00fcetleri kullan\u0131larak videodan otomatik taranmas\u0131. Sa\u011fdaki g\u00f6rsel solda oynamakta olan videodan hareketli alanlar\u0131 g\u00f6stermektedir. Alttaki grafik zaman i\u00e7indeki hareket miktar\u0131n\u0131 g\u00f6sterir.<\/i><\/span><\/p>\n

Y\u00fcz temelli duygu alg\u0131lay\u0131c\u0131lar ile ilgili bir s\u0131n\u0131rl\u0131l\u0131k, yaln\u0131zca y\u00fcz video ak\u0131\u015f\u0131nda otomatik olarak saptanabildi\u011finde uygulanabilir olmalar\u0131d\u0131r. Bu a\u015f\u0131r\u0131 hareket, kapanma, az \u0131\u015f\u0131k ve di\u011fer fakt\u00f6rlere ba\u011fl\u0131 olarak her zaman m\u00fcmk\u00fcn olmamaktad\u0131r. Asl\u0131nda y\u00fcz-temelli duygu saptay\u0131c\u0131lar t\u00fcm durumlar\u0131n %65’ine uygulanabilmektedir. Buna de\u011finmek i\u00e7in Bosch, Chen, Baker, Shute ve D’Mello (2015) \u00e7okbi\u00e7imli teknikleri etkile\u015fim temelli (\u00f6nceki b\u00f6l\u00fcme benzer olarak) ve y\u00fcz-temelli saptay\u0131c\u0131larla birle\u015ftirmek i\u00e7in kulland\u0131lar. Etkile\u015fim tabanl\u0131 saptay\u0131c\u0131lar, y\u00fcz bazl\u0131 saptay\u0131c\u0131lardan daha az do\u011fruydu (Kai vd., 2015) ancak neredeyse t\u00fcm vakalara uygulanabilirdi. Bu ikisinin birle\u015fimi ile saptay\u0131c\u0131lar\u0131n durumlara uygulanabilirli\u011fi y\u00fcz temelli saptay\u0131c\u0131lara k\u0131yasla kesinlikte k\u00fc\u00e7\u00fck bir azalma ile (<%5 farkla) %98’e y\u00fckseltilmi\u015f oldu.<\/p>\n

Duygu Modellerini Duygu-Bilir \u00d6\u011frenme Teknolojilerine Entegre Etmek<\/h3>\n

Yukar\u0131da tart\u0131\u015f\u0131lan etkile\u015fim ve bedensel temelli etki alg\u0131lay\u0131c\u0131lar\u0131, bir \u00f6\u011frenme teknolojisi ile etkile\u015fimler s\u0131ras\u0131nda \u00f6\u011frenenin etkilerinin ger\u00e7ek zamanl\u0131 de\u011ferlendirmelerini sa\u011flamak i\u00e7in kullan\u0131labilecek somut eserlerdir. Bu durum hissedilen duyguya dinamik anlamda cevap vererek heyecan verici bir d\u00f6ng\u00fcy\u00fc kapatma olas\u0131l\u0131\u011f\u0131n\u0131 destekler. Bu gibi duygu-bilir \u00f6\u011frenme teknolojilerinin amac\u0131, \u00f6\u011frencilerin ne hissettiklerine ek olarak ne d\u00fc\u015f\u00fcnd\u00fckleri ya da yapt\u0131klar\u0131na cevap vererek g\u00fcncel \u00f6\u011frenme teknolojilerinin uyarlanabilirlik bant aral\u0131\u011f\u0131n\u0131n geni\u015fletmektir. (\u0130nceleme i\u00e7in, bk. D\u2019Mello, Blanchard, Baker, Ocumpaugh ve Brawner, 2014). Burada ben, b\u00f6ylesine iki sisteme, Affective AutoTutor (D’Mello ve Graesser, 2012a) ve UNC-ITSPOKE (Forbes-Riley ve Litman, 2011) dikkat \u00e7ekiyorum.<\/p>\n

\"\"<\/p>\n

\u015eekil 10.4. Affective AutoTutor: \u00d6\u011frencilerin s\u0131k\u0131nt\u0131lar\u0131n\u0131, kafa kar\u0131\u015f\u0131kl\u0131klar\u0131n\u0131 ve s\u0131k\u0131nt\u0131lar\u0131n\u0131 otomatik olarak alg\u0131layan ve bunlara yan\u0131t veren konu\u015fma diyaloglar\u0131na sahip ak\u0131ll\u0131 \u00f6\u011fretim sistemi (A\u00d6S).<\/i><\/span><\/p>\n

Affective AutoTutor (bk. \u015eekil 10.4) do\u011fal dil i\u00e7inde karma-inisiyatif bir diyalog d\u00fczenleyerek \u00f6\u011frencilerin Newton fizi\u011fi, bilgisayar okur yazarl\u0131\u011f\u0131 ve bilimsel ak\u0131l y\u00fcr\u00fctme gibi zor konularda uzmanl\u0131k geli\u015ftirmelerine yard\u0131m eden konu\u015fma tabanl\u0131 bir A\u00d6S olan AutoTutor’un de\u011fi\u015ftirilmi\u015f bir s\u00fcr\u00fcm\u00fcd\u00fcr. (Graesser, Chipman, Haynes ve Olney, 2005 ). \u00d6zg\u00fcn AutoTutor sistemi \u00f6\u011frenenin bili\u015fsel durumlar\u0131na kar\u015f\u0131 duyarl\u0131 olan bir dizi belirsiz \u00fcretim kurallar\u0131na sahipti. Affective AutoTutor bu kurallar\u0131 \u00f6\u011frenenlerin duyu\u015fsal durumlar\u0131 \u00f6zellikle can s\u0131k\u0131nt\u0131s\u0131, kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131 ve hayal k\u0131r\u0131kl\u0131\u011f\u0131n\u0131n dinamik olarak de\u011ferlendirilmesine duyarl\u0131 olacak \u015fekilde artt\u0131r\u0131r. Duyu\u015fsal durumlar etkile\u015fim \u00f6r\u00fcnt\u00fclerin, b\u00fcy\u00fck beden hareketlerini ve y\u00fcz niteliklerini otomatik olarak izleyerek hissedilir. (D’Mello ve Graesser, 2012a). Affective AutoTutor duygusal g\u00f6r\u00fcn\u00fcmlerin yan\u0131 s\u0131ra empatik, y\u00fcreklendirici ve motive edici diyalog-hareketleri ile cevap verir. \u00d6rne\u011fin, asistan hafif can s\u0131k\u0131nt\u0131s\u0131na, “Bunlar bazen biraz s\u0131k\u0131c\u0131 olabiliyor, bu y\u00fczden senin bu i\u015fin \u00fcstesinden gelmene yard\u0131mc\u0131 olmay\u0131 deneyece\u011fim. Hadi gidelim”. Duyu\u015fsal cevaplara uygun duygusal y\u00fcz ifadeleri ve duygusal olarak d\u00fczenlenmi\u015f konu\u015fmalar (\u00f6r. sentezlenmi\u015f duyguda\u015fl\u0131k veya y\u00fcreklendirme) e\u015flik eder.<\/p>\n

Affective AutoTutor’un \u00f6zg\u00fcn duyu\u015fsal olmayan AutoTutora g\u00f6re etkilili\u011fi 84 \u00f6\u011frenenin rastgele olarak iki\u015fer 30 dakikal\u0131k \u00f6\u011frenme oturumuna atand\u0131\u011f\u0131 denekler aras\u0131 desenle test edilmi\u015ftir. (D’Mello, Lehman, Sullins vd., 2010). Sonu\u00e7lar duyu\u015fsal asistan\u0131n ikinci 30 dakikal\u0131k \u00f6\u011frenme oturumunda d\u00fc\u015f\u00fck bilgi alan\u0131 \u00f6\u011frenenlerinin \u00f6\u011frenmesinde yard\u0131mc\u0131 oldu\u011funu g\u00f6stermi\u015ftir. Duyu\u015fsal asistan ilk 30 dakikal\u0131k oturumda \u00fcst bilgi alan\u0131 \u00f6\u011frenenlerinin \u00f6\u011frenmesini desteklemekte daha az etkili olmu\u015ftur. \u00d6nemli bi\u00e7imde, duyu\u015fsal asistanla \u00f6\u011frenme kazan\u0131mlar\u0131 Oturum 1’den Oturum 2’ye y\u00fckselmi\u015f oysaki duyu\u015fsal olmayan asistanla y\u00fckseli\u015f sonras\u0131 dura\u011fan bir noktaya gelmi\u015ftir. Duyu\u015fsal asistanla etkile\u015fim kuran \u00f6\u011frenenler bir sonraki transfer s\u0131nav\u0131nda \u00e7ok daha iyi bir performans g\u00f6sterdiler. Takip eden analiz g\u00f6sterdi ki, \u00f6\u011frenenlerin \u00f6\u011frenme oturumlar\u0131 boyunca bilgisayar \u00f6\u011fretenlerin (asistanlar\u0131n) insan \u00f6\u011fretenlere ne kadar benzedi\u011fine dair alg\u0131lar\u0131ndaki art\u0131\u015f asistan geri bildirim niteli\u011fine ba\u011fl\u0131d\u0131r ve \u00f6\u011frenmenin g\u00fc\u00e7l\u00fc bir yorday\u0131c\u0131s\u0131d\u0131r. (D’Mello ve Graesser, 2012c). Duyu\u015fsal asistan i\u00e7in alg\u0131daki olumlu de\u011fi\u015fim daha fazla idi.<\/p>\n

\u0130kinci bir \u00f6rnek olarak, \u00f6\u011frenenlerin s\u00f6zl\u00fc cevaplar\u0131n\u0131n kesin\/kesin olmamas\u0131 ve do\u011fru\/do\u011fru olmamas\u0131n\u0131 otomatik olarak saptama ve cevap verme becerisine sahip olan konu\u015fma yetene\u011fi etkinle\u015ftirilmi\u015f bir fizik A\u00d6S’i olan UNC-ITSPKOE (Forbes-Riley ve Litman, 2011) \u00f6rne\u011fini inceleyelim. Kesin olmaman\u0131n saptanmas\u0131 \u00f6\u011frenenlerin s\u00f6zl\u00fc cevaplar\u0131n\u0131n s\u00f6zc\u00fcksel ve diyalog tabanl\u0131 niteliklerinin yan\u0131 s\u0131ra akustik-prosodik niteliklerinin ay\u0131klanmas\u0131 ve analiz edilmesi ile ger\u00e7ekle\u015ftirilir. UNC-ITSPOKE \u00f6\u011frenenin cevab\u0131n\u0131n do\u011fru oldu\u011fu fakat emin olmad\u0131\u011f\u0131 durumlarda kesin olmama durumuna cevap verdi. Bu i\u00e7inden \u00e7\u0131k\u0131lmaz g\u00fc\u00e7 bir durumun g\u00f6stergesi olarak al\u0131nd\u0131 \u00e7\u00fcnk\u00fc \u00f6\u011frenen do\u011fru olmas\u0131na ra\u011fmen bilgisinin durumundan emin de\u011fildi. Mevcut cevap stratejisi belirsizli\u011fi \u00e7\u00f6zecek bir ek e\u011fitim sa\u011flayan a\u00e7\u0131klama temelli yan diyaloglar\u0131n faaliyete ge\u00e7irilmesini i\u00e7ermekteydi. Bu da ilaveten takip eden sorular\u0131 (daha zor i\u00e7erik i\u00e7in) veya basit\u00e7e do\u011fru bilginin ayr\u0131nt\u0131l\u0131 a\u00e7\u0131klamalarla (daha kolay i\u00e7erik i\u00e7in) tasdik edilmesini \u0130\u00e7erebilirdi.<\/p>\n

Forbes-Riley ve Litman (2011) kesin olmama durumunda uyarlanabilir cevaplar\u0131 alma (uyarlanabilir durum), kesin olmama durumunda hi\u00e7 cevap almama (uyarlanamayan durum) veya kesin olmama durumunda rastgele cevaplar alma (rastgele kontrol durumu) i\u00e7in rastgele atanan 72 tane \u00f6\u011frenenin \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131n\u0131 kar\u015f\u0131la\u015ft\u0131rd\u0131lar. Bu son durumda, yan diyaloglardan eklenen \u00f6\u011fretim i\u00e7eri\u011fi ilave \u00f6\u011fretimi kontrol etmek ad\u0131na rastgele bir dizi de\u011fi\u015fiklik i\u00e7in verilmi\u015fti. Bulgular uyarlanabilir durumun rastgele ve uyarlanamayan kontrol durumlar\u0131na nazaran az miktarda (fakat manidar olmayan d\u00fczeyde)ileri \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131 elde etti\u011fini g\u00f6sterdi. Bulgular \u00f6\u011frenme \u00e7\u0131kt\u0131lar\u0131 ile ili\u015fkili olan\u0131n belki de kesin olmama durumunda uyarlanabilir cevaplar\u0131n varl\u0131\u011f\u0131 ya da yoklu\u011funun de\u011fil fakat uyarlanabilir cevaplar\u0131n say\u0131s\u0131n\u0131n oldu\u011funu a\u00e7\u0131\u011fa \u00e7\u0131kard\u0131.<\/p>\n

GEL\u0130\u015eMEKTE OLAN TEMALAR<\/h2>\n

Duygular, \u00f6\u011frenme, \u00d6A ve EVM’nin kesi\u015fimindeki ara\u015ft\u0131rmalar genel olarak bilgisayar destekli ak\u0131ll\u0131 \u00f6\u011fretim sistemleriyle bire bir \u00f6\u011frenmeye (Forbes-Riley ve Litman, 2011; Woolf vd., 2009), e\u011fitsel oyunlara (Conati ve Maclaren, 2009; Sabourin, Mott ve Lester 2011) veya okuma, yazma, metin-diyagram entegrasyonu ve problem \u00e7\u00f6zme gibi temel yeterlilikleri destekleyen aray\u00fczler (D’Mello ve Graesser, 2014a; D’Mello, Lehman ve Person, 2010; D’Mello ve Mills, 2014). Bu ana ara\u015ft\u0131rma kollar\u0131 olduk\u00e7a aktif olmas\u0131na ra\u011fmen, son g\u00fcnlerdeki ara\u015ft\u0131rmalar duygunun, \u00f6\u011frenmeyi kapsayan daha geni\u015f sosyok\u00fclt\u00fcrel ba\u011flam\u0131 daha yak\u0131ndan yans\u0131tacak \u00e7ok daha kapsay\u0131c\u0131 etkile\u015fim ba\u011flamlar\u0131 boyunca analizine odaklanm\u0131\u015ft\u0131r. Ben baz\u0131 heyecan verici geli\u015fmeleri \u00f6rneklendirmek a\u00e7\u0131\u015f\u0131ndan k\u0131saca d\u00f6rt ara\u015ft\u0131rma temas\u0131 tan\u0131mlayaca\u011f\u0131m.<\/p>\n

Y\u0131pranma ve Okul Terkinin Duygu Temelli Yorday\u0131c\u0131lar\u0131<\/h3>\n

Erken risk g\u00f6stergeleri ve erken m\u00fcdahale sistemleri \u00d6A ve EVM’nin “muhte\u015fem uygulamalar\u01312<\/sup><\/a>” d\u0131r. (Jayaprakash, Moody, Lauria, Regan ve Baron, 2014). Alandaki \u00e7o\u011fu sistemler akademik performans verisi, demografik veriler ve finansal yard\u0131m\u0131n ula\u015f\u0131labilirli\u011fi \u00fczerine odaklanm\u0131\u015ft\u0131r. Bu fakt\u00f6rler \u015f\u00fcphesiz \u00f6nemlidir fakat devreye girmesi muhtemel di\u011fer de\u011fi\u015fimli fakt\u00f6rler vard\u0131r. Bunu ak\u0131lda tutarak, Aguiar, Ambrose, Chawla, Goodrich ve Brockman (2014) bir m\u00fchendisli\u011fe giri\u015f dersini b\u0131rakma durumlar\u0131n\u0131 yordamada geleneksel akademik ve demografik niteliklerin yordama g\u00fcc\u00fcn\u00fc davran\u0131\u015fsal kat\u0131l\u0131m\u0131 g\u00f6steren niteliklerle kar\u015f\u0131la\u015ft\u0131rd\u0131lar. Temel bulgular\u0131, oturum a\u00e7ma say\u0131lar\u0131, g\u00f6nderilen \u00e7al\u0131\u015fmalar\u0131n say\u0131s\u0131 ve sayfa t\u0131klanma say\u0131lar\u0131yla \u00f6l\u00e7\u00fclen e-portfolyolarla davran\u0131\u015fsal olarak me\u015fguliyet durumunun ders b\u0131rakmay\u0131, sadece akademik performans ve demografik yap\u0131lardan olu\u015fturulan modellerden daha iyi yordayabildi\u011fiydi. Bu \u00e7al\u0131\u015fmada duygu do\u011frudan olarak \u00f6l\u00e7\u00fclmemi\u015f olsa da e portfolyolara davran\u0131\u015fsal olarak kat\u0131l\u0131m, g\u00fc\u00e7l\u00fc bir g\u00fcd\u00fcleyici duygu olan ilginin bir i\u015fareti olarak d\u00fc\u015f\u00fcn\u00fclebilir.<\/p>\n

Tart\u0131\u015fma Forumlar\u0131n\u0131n Duygu Analizleri<\/h3>\n

Dil duygular\u0131 ileti\u015fime ge\u00e7irir. Dolay\u0131s\u0131yla duygu analizi ve fikir madencili\u011fi teknikleri (Pang ve Lee, 2008) \u00f6\u011frencileri bir \u00f6\u011frenme deneyimi hakk\u0131ndaki d\u00fc\u015f\u00fcncelerinin (yaz\u0131l\u0131 dilde ifade edilen) ili\u015fkili davran\u0131\u015flar\u0131 (\u00f6zellikle y\u0131pranma) nas\u0131l yordad\u0131\u011f\u0131n\u0131 \u00e7al\u0131\u015f\u0131rken \u00f6nemli d\u00fczeyde bir g\u00fcce sahiptir. Bu do\u011frultuda Wen, Yang ve Rose (2014) \u00fc\u00e7 tane kitlesel a\u00e7\u0131k \u00e7evrimi\u00e7i derslerdeki (KA\u00c7D) \u00f6\u011frenen g\u00f6nderilerine duygu analizi tekniklerini uygulad\u0131lar. Olumlunun olumsuz terimlere oran\u0131 ile okul b\u0131rakman\u0131n zamana oran\u0131 aras\u0131nda negatif ili\u015fki g\u00f6zlemlediler. Daha da yak\u0131n d\u00f6nemde, Yang, Wen, Howley, Kraut ve Rose (2015) \u00f6\u011frencideki kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n belirleyicisi olan tart\u0131\u015fma g\u00f6nderilerini otomatik olarak belirleyecek metotlar geli\u015ftirdiler. Kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n okulu s\u00fcrd\u00fcrebilme olas\u0131l\u0131\u011f\u0131n\u0131 d\u00fc\u015f\u00fcrd\u00fc\u011f\u00fcn\u00fc, fakat bunun kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131 \u00e7\u00f6zme ve di\u011fer destekleyici m\u00fcdahalelerle hafifletilebilece\u011fini g\u00f6sterdiler.<\/p>\n

S\u0131n\u0131f \u00d6\u011frenme Analitikleri<\/h3>\n

Alg\u0131lama ve sinyal i\u015fleme teknolojilerindeki son geli\u015fmeler daha \u00f6nceden sadece ki\u015fisel raporlar ve kullan\u0131\u015fs\u0131z insan g\u00f6zlemleri ile elde edilebilecek olan \u00f6\u011frencilerin s\u0131n\u0131f deneyimlerine ili\u015fkin halleri otomatik olarak modelleyebilmeyi sa\u011flad\u0131. \u00d6rne\u011fin, ikinci ku\u015fak Kinects g\u00f6zlerin ya da a\u011fz\u0131n a\u00e7\u0131k olup olmad\u0131\u011f\u0131n\u0131, ki\u015finin ba\u015fka bir yere bak\u0131p bakmad\u0131\u011f\u0131n\u0131 ya da a\u011fz\u0131n\u0131n oynay\u0131p oynamad\u0131\u011f\u0131n\u0131 bir defada alt\u0131 ki\u015fiye kadar tespit edebilmektedir. (Microsoft, 2015). \u00d6nc\u00fc bir \u00e7al\u0131\u015fmada, Raca, Kidzinski ve Dillenbourg (2015), \u00f6\u011frencileri karatahta alan\u0131n\u0131n etraf\u0131na yerle\u015ftirilmi\u015f birden fazla kamera kullanarak s\u0131n\u0131fta izlemi\u015ftir. Daha sonra bir saptay\u0131c\u0131y\u0131 e\u011fitmek i\u00e7in kullan\u0131lan kafa alg\u0131lama ve kafa pozu tahmini i\u00e7in bilgisayarl\u0131 g\u00f6rme teknikleri kullan\u0131lm\u0131\u015ft\u0131r. \u00d6\u011frencilerin dikkatini \u00e7ekme (kendi kendine raporlama yoluyla do\u011fruland\u0131). \u00c7ok modlu \u00f6\u011frenme analiti\u011fi alan\u0131yla (Blikstein, 2013) ili\u015fkili olan geli\u015fmekteki bu alan, \u00f6n\u00fcm\u00fczdeki uzun y\u0131llar i\u00e7inde \u00f6nemli geli\u015fmelere haz\u0131rd\u0131r.<\/p>\n

\u00d6\u011fretmen Analitikleri<\/h3>\n

\u00d6\u011fretmen uygulamalar\u0131n\u0131n \u00f6\u011frenci duygu ve kat\u0131l\u0131mlar\u0131n\u0131 etkiledi\u011fi bilindi\u011finden \u00f6\u011fretmenler d\u00f6ng\u00fcn\u00fcn d\u0131\u015f\u0131nda b\u0131rak\u0131lmamal\u0131d\u0131r. Ne yaz\u0131k ki, \u00f6\u011fretmenlerin \u00f6\u011fretim uygulamalar\u0131n\u0131n niceli\u011fi s\u0131n\u0131flardaki canl\u0131 g\u00f6zlemlere dayan\u0131r (\u00f6r. Nystrand, 1997) ve bu da ara\u015ft\u0131rman\u0131n \u00f6l\u00e7eklendirilmesini zorla\u015ft\u0131r\u0131r. Bunu ele almak i\u00e7in, ara\u015ft\u0131rmac\u0131lar \u00f6\u011fretmen \u00f6\u011fretimsel uygulamalar\u0131n\u0131n otomatik analizi i\u00e7in y\u00f6ntemler geli\u015ftirmeye ba\u015flad\u0131lar. \u00d6nc\u00fc olan \u00e7al\u0131\u015fmalardan birinde, Wang, Miller ve Cortina (2013) 1 ve 3’\u00fcnc\u00fc s\u0131n\u0131f matematik derslerinin ses kay\u0131tlar\u0131n\u0131 alm\u0131\u015flar ve bu derslerdeki tart\u0131\u015fmalar\u0131n d\u00fczeylerinin belirleyecek otomatik metotlar geli\u015ftirmi\u015flerdir. Bu \u00e7al\u0131\u015fma yak\u0131n zamanda daha b\u00fcy\u00fck orta \u00f6\u011fretim alan yaz\u0131n \u00f6rneklemeleri ve sadece \u00f6\u011fretmen sesi kullan\u0131lan dil s\u0131n\u0131f\u0131 \u00f6rneklemelerinde bir \u00e7ok ek \u00f6\u011fretim etkinli\u011fi (ders anlatma, k\u00fc\u00e7\u00fck grup \u00e7al\u0131\u015fmas\u0131, denetimli s\u0131ra \u00e7al\u0131\u015fmas\u0131, soru\/cevap ve y\u00f6nergeler ve y\u00f6nler) (Donnelly vd., 2016a) veya \u00f6\u011fretmen ve s\u0131n\u0131f ses kayd\u0131n\u0131n birle\u015ftirilmesini (Donnelly vd., 2016b) analiz etmek amac\u0131yla geni\u015fletilmi\u015ftir. Blanchard vd. (2016) insanlar taraf\u0131ndan kodlanan sorular ile .85 korelasyon sa\u011flayarak, \u00f6\u011fretmen sorular\u0131n\u0131 otomatik olarak saptayabilmek i\u00e7in \u00f6\u011fretmen ses kay\u0131tlar\u0131n\u0131 kulland\u0131. Bu alandaki \u00e7al\u0131\u015fmalarda bir sonraki a\u015fama ise, \u00f6\u011frencilerin nas\u0131l hissettiklerini bir de ne d\u00fc\u015f\u00fcnd\u00fckleri, yapt\u0131klar\u0131 ve \u00f6\u011frendiklerini etkileyen \u00e7evredeki di\u011fer \u00f6gelerle birlikte de\u011ferlendirebilmek i\u00e7in \u00f6\u011fretmenlerin ne yapt\u0131klar\u0131na dair bilgiyi kullanmak olacakt\u0131r.<\/p>\n

GELECEK TEMALARI<\/h2>\n

Ara\u015ft\u0131rmaya dair baz\u0131 olas\u0131 gelecek temalar\u0131n\u0131 k\u0131saca vurgulayarak bitireyim. Umut verici ara\u015ft\u0131rma alanlar\u0131ndan biri \u00f6\u011frenenlerin ve \u00f6\u011frenme topluluklar\u0131n\u0131n duygusal deneyimlerinin geleneksel bir s\u0131n\u0131f\u0131n, ters y\u00fcz edilmi\u015f s\u0131n\u0131f\u0131n, ya da KA\u00c7D ‘\u0131n geni\u015fletilmi\u015f zaman \u00f6l\u00e7e\u011finde detayl\u0131 bir analizini i\u00e7ermektedir. (Dillon vd., 2016). \u0130kincisi faydal\u0131 olanlar meydana \u00e7\u0131kar\u0131labilsin diye (\u00f6r. Strain ve D’Mello, 2014), \u00f6\u011frenme esnas\u0131ndaki duygu d\u00fczenlemenin, \u00f6zellikle \u00d6A\/EVM metotlar\u0131n\u0131n farkl\u0131 d\u00fczenleme stratejilerini belirlemede nas\u0131l kullan\u0131labilece\u011finin \u00e7al\u0131\u015f\u0131lmas\u0131d\u0131r (Gross, 2008). \u00dc\u00e7\u00fcnc\u00fcs\u00fc duygunun yan\u0131 s\u0131ra fark\u0131ndal\u0131k dikkat durumlar\u0131, konu ile ili\u015fkili olmayan durumlar\u0131n d\u00fc\u015f\u00fcn\u00fclmesi ve duygu- dikkat kar\u0131\u015f\u0131m\u0131n\u0131n “ak\u0131\u015f-deneyimi” ne benzer \u015fekilde nas\u0131l harmanland\u0131\u011f\u0131n\u0131n (Csikszentmihalyi, 1990) ortaya \u00e7\u0131k\u0131\u015f\u0131 ve beden ve davran\u0131\u015fta g\u00f6r\u00fcn\u00fcr olu\u015funu birlikte d\u00fc\u015f\u00fcnecektir. D\u00f6rd\u00fcnc\u00fcs\u00fc s\u00f6zde “bili\u015fsel olmayan” (Farrington vd., 2012) azim, oto kontrol ve gayret gibi ki\u015fisel \u00f6zelliklerin \u00f6\u011frenen duygular\u0131n\u0131 ve onlar\u0131 d\u00fczenleme \u00e7abalar\u0131n\u0131 nas\u0131l denetledi\u011fine de\u011finir (\u00f6r. Galla vd., 2014). Be\u015fincisi i\u015f birli\u011fi yapman\u0131n kritik bir 21. y\u00fczy\u0131l becerisi olarak \u00f6nemine istinaden (OECD, 2015), i\u015fbirlikli \u00f6\u011frenme ve problem \u00e7\u00f6zme esnas\u0131nda \u00f6\u011frenen gruplar\u0131n\u0131n duygular\u0131n\u0131 izleyebilir (Ringeval, Sonderegger, Sauer ve Lalanne, 2013).<\/p>\n

Son olarak, William James’in 1884 tarihli duygu \u00fczerine klasik tezinden al\u0131nt\u0131 yaparak “\u00c7evremin par\u00e7alar\u0131n\u0131n en \u00f6nemlisi benim adam\u0131md\u0131r. Onun bana olan tutumunun bilinci; utan\u00e7lar\u0131m\u0131n, k\u0131zg\u0131nl\u0131klar\u0131m\u0131n ve korkular\u0131m\u0131n \u00e7o\u011funun normal olarak \u00e7\u00f6zecek olan alg\u0131d\u0131r” (s. 195). G\u00fcn\u00fcm\u00fcze kadar olan ara\u015ft\u0131rmalar temelde ba\u015far\u0131, bilgiye dair ve konuya dair duygulara odaklanm\u0131\u015ft\u0131r. Ancak \u00f6\u011frenmenin ger\u00e7ekle\u015fti\u011fi sosyo k\u00fclt\u00fcrel ba\u011flam\u0131n analizi mutlaka gurur, su\u00e7luluk, k\u0131skan\u00e7l\u0131k, haset gibi sosyal duygulara duygun bir \u015fekilde de\u011finmelidir. Bu hem bir gelecek temas\u0131 ve hem de b\u00fcy\u00fck bir ara\u015ft\u0131rma meydan okumas\u0131d\u0131r.<\/p>\n

SONU\u00c7<\/h2>\n

\u00d6\u011frenme so\u011fuk bir entelekt\u00fcel etkinlik de\u011fildir; duygularla i\u015faretlenmi\u015ftir. Duygular sadece ifadelerin s\u00fcs\u00fc de\u011fildir ayn\u0131 zamanda temsil g\u00f6revleri de vard\u0131r. Fakat duygu \u00e7oklu zaman \u00f6l\u00e7eklerinde devingen olarak geli\u015fen \u00e7oklu bile\u015fenlere sahip karma\u015f\u0131k bir fenomendir. Duyu\u015fsal bilimler ve duyu\u015fsal sinir bilimdeki b\u00fcy\u00fck ad\u0131mlara ra\u011fmen duygular hakk\u0131nda \u00e7ok az ve hatta \u00f6\u011frenme esnas\u0131ndaki duygular hakk\u0131nda daha da az \u015fey biliyoruz. Bu kesinlikle teorik olarak bir belirginlik olu\u015fana kadar duygular\u0131 modellemekten ka\u00e7\u0131nmam\u0131z gerekti\u011fi anlam\u0131na gelmez. Ger\u00e7ekte durum tam tersidir. Basit\u00e7e duygular\u0131 modelledi\u011fimizi s\u00f6yledi\u011fimizde neyi modelledi\u011fimiz konusunda daha dikkatli olmal\u0131y\u0131z anlam\u0131na gelir. Ayn\u0131 zamanda duygunun i\u00e7indeki karma\u015f\u0131kl\u0131\u011f\u0131 ve belirsizli\u011fin etkisini azaltmak yerine onlar\u0131 benimsemeliyiz. Bilakis, bulgu temelli, veriye dayal\u0131, \u00d6A ve EVM’nin analitik y\u00f6ntemlerinden biri ger\u00e7ek ya\u015fam veri toplaman\u0131n beraberinde hem \u00f6\u011frenme bilimi hem de duygu bilimine geli\u015ftirecek benzersiz potansiyele sahiptir. Her \u015fey \u00f6\u011frenme analizine duygular\u0131 dahil ederek ba\u015flar.<\/p>\n

TE\u015eEKK\u00dcR B\u00d6L\u00dcM\u00dc<\/h2>\n

Bu ara\u015ft\u0131rma Ulusal Bilim Vakf\u0131 (UBV) (DRL 1108845 ve IIS 1523091), Bill ve Melinda Gates Vakf\u0131 ve E\u011fitim Bilimleri Enstit\u00fcs\u00fc (R305A130030) taraf\u0131ndan desteklenmi\u015ftir. Bu makalede ifade edilen d\u00fc\u015f\u00fcnceler, bulgular ve \u00e7\u0131kar\u0131mlar veya tavsiyeler yazara ait olup, destek veren kurulu\u015flar\u0131n g\u00f6r\u00fc\u015flerini yans\u0131tmas\u0131 gerekmemektedir.<\/p>\n

KAYNAK\u00c7A<\/h2>\n

Aguiar, E., Ambrose, G. A. A., Chawla, N. V., Goodrich, V., & Brockman, J. (2014). Engagement vs. performance: Using electronic portfolios to predict first semester engineering student persistence. Journal of Learning Analytics, 1<\/i>(3), 7\u201333. <\/span><\/p>\n

Ai, H., Litman, D. J., Forbes-Riley, K., Rotaru, M., Tetreault, J. R., & Pur, A. (2006). Using system and user performance features to improve emotion detection in spoken tutoring dialogs. Proceedings of the 9th International Conference on Spoken Language Processing <\/i>(Interspeech 2006). <\/span><\/p>\n

Arroyo, I., Woolf, B., Cooper, D., Burleson, W., Muldner, K., & Christopherson, R. (2009). Emotion sensors go to school. In V. Dimitrova, R. Mizoguchi, B. Du Boulay, & A. Graesser (Eds.), Proceedings of the 14th International Conference on Artificial Intelligence in Education <\/i>(pp. 17\u201324). Amsterdam: IOS Press. <\/span><\/p>\n

Baker, R., & Ocumpaugh, J. (2015). Interaction-based affect detection in educational software. In R. Calvo, S. D’Mello, J. Gratch, & A. Kappas (Eds.), The Oxford handbook of affective computing <\/i>(pp. 233\u2013245). New York: Oxford University Press. <\/span><\/p>\n

Barth, C. M., & Funke, J. (2010). Negative affective environments improve complex solving performance. Cognition and Emotion, 24<\/i>(7), 1259\u20131268. doi:10.1080\/02699930903223766 <\/span><\/p>\n

Blanchard, N., Donnelly, P., Olney, A. M., Samei, B., Ward, B., Sun, X., . . . D’Mello, S. K. (2016). Identifying teacher questions using automatic speech recognition in live classrooms. Proceedings of the 17th Annual SIGdial Meeting on Discourse and Dialogue <\/i>(SIGDIAL 2016) (pp. 191\u2013201). Association for Computational Linguistics. <\/span><\/p>\n

Blikstein, P. (2013). Multimodal learning analytics. Proceedings of the 3rd International Conference on Learning Analytics and Knowledge <\/i>(LAK\u201913), 8\u201312 April 2013, Leuven, Belgium. New York: ACM. <\/span><\/p>\n

Bosch, N., Chen, H., Baker, R., Shute, V., & D’Mello, S. K. (2015). Accuracy vs. availability heuristic in multimodal affect detection in the wild. Proceedings of the 17th ACM International Conference on Multimodal Interaction <\/i>(ICMI 2015). New York: ACM. <\/span><\/p>\n

Bosch, N., & D’Mello, S. K. (in press). The affective experience of novice computer programmers. International Journal of Artificial Intelligence in Education<\/i>. <\/span><\/p>\n

Bosch, N., D’Mello, S., Baker, R., Ocumpaugh, J., & Shute, V. (2016). Using video to automatically detect learner affect in computer-enabled classrooms. ACM Transactions on Interactive Intelligent Systems, 6<\/i>(2), doi:10.1145\/2946837. <\/span><\/p>\n

Calvo, R. A., & D’Mello, S. K. (2010). Affect detection: An interdisciplinary review of models, methods, and their applications. IEEE Transactions on Affective Computing, 1<\/i>(1), 18\u201337. doi:10.1109\/T-AFFC.2010.1 <\/span><\/p>\n

Clore, G. L., & Huntsinger, J. R. (2007). How emotions inform judgment and regulate thought. Trends in Cognitive Sciences, 11<\/i>(9), 393\u2013399. doi: 10.1016 \/ j.tics.2007.08.005 <\/span><\/p>\n

Conati, C., & Maclaren, H. (2009). Empirically building and evaluating a probabilistic model of user affect. User Modeling and User-Adapted Interaction, 19<\/i>(3), 267\u2013303. <\/span><\/p>\n

Corbett, A., & Anderson, J. (1995). Knowledge tracing: Modeling the acquisition of procedural knowledge. User Modeling and User-Adapted Interaction, 4<\/i>(4), 253\u2013278. <\/span><\/p>\n

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience<\/i>. New York: Harper & Row. <\/span><\/p>\n

D’Mello, S. K. (2016). On the influence of an iterative affect annotation approach on inter-observer and self-observer reliability. IEEE Transactions on Affective Computing, 7<\/i>(2), 136\u2013149. <\/span><\/p>\n

D’Mello, S. K., Blanchard, N., Baker, R., Ocumpaugh, J., & Brawner, K. (2014). I feel your pain: A selective review of affect-sensitive instructional strategies. In R. Sottilare, A. Graesser, X. Hu, & B. Goldberg (Eds.), Design recommendations for adaptive intelligent tutoring systems: Adaptive instructional strategies <\/i>(Vol. 2, pp. 35\u201348). Orlando, FL: US Army Research Laboratory. <\/span><\/p>\n

D’Mello, S., Craig, S., Sullins, J., & Graesser, A. (2006). Predicting affective states expressed through an emote-aloud procedure from AutoTutor’s mixed-initiative dialogue. International Journal of Artificial Intelligence in Education, 16<\/i>(1), 3\u201328.<\/span><\/p>\n

D’Mello, S., & Graesser, A. (2012a). AutoTutor and Affective AutoTutor: Learning by talking with cognitively and emotionally intelligent computers that talk back. ACM Transactions on Interactive Intelligent Systems, 2<\/i>(4), 23:22\u201323:39. <\/span><\/p>\n

D’Mello, S., & Graesser, A. (2012b). Dynamics of affective states during complex learning. Learning and Instruction, 22<\/i>(2), 145\u2013157. doi: 10.1016 \/ j.learninstruc.2011.10.001 <\/span><\/p>\n

D’Mello, S., & Graesser, A. (2012c). Malleability of students’ perceptions of an affect-sensitive tutor and its influence on learning. In G. Youngblood & P. McCarthy (Eds.), Proceedings of the 25th Florida Artificial Intelligence Research Society Conference <\/i>(pp. 432\u2013437). Menlo Park, CA: AAAI Press. <\/span><\/p>\n

D’Mello, S., & Graesser, A. (2014a). Inducing and tracking confusion and cognitive disequilibrium with breakdown scenarios. Acta Psychologica, 151<\/i>, 106\u2013116. <\/span><\/p>\n

D’Mello, S. K., & Graesser, A. C. (2014b). Confusion. In R. Pekrun & L. Linnenbrink-Garcia (Eds.), International handbook of emotions in education <\/i>(pp. 289\u2013310). New York: Routledge. <\/span><\/p>\n

D’Mello, S. K., & Kory, J. (2015). A review and meta-analysis of multimodal affect detection systems. ACM Computing Surveys, 47<\/i>(3), 43:41\u201343:46. <\/span><\/p>\n

D’Mello, S., Lehman, B., & Person, N. (2010). Monitoring affect states during effortful problem solving activities. International Journal of Artificial Intelligence in Education, 20<\/i>(4), 361\u2013389. <\/span><\/p>\n

D’Mello, S., Lehman, B., Sullins, J., Daigle, R., Combs, R., Vogt, K., . . . Graesser, A. (2010). A time for emoting: When affect-sensitivity is and isn’t effective at promoting deep learning. In J. Kay & V. Aleven (Eds.), Proceedings of the 10th International Conference on Intelligent Tutoring Systems <\/i>(pp. 245\u2013254). Berlin \/ Heidelberg: Springer. <\/span><\/p>\n

D’Mello, S., & Mills, C. (2014). Emotions while writing about emotional and non-emotional topics. Motivation and Emotion, 38<\/i>(1), 140\u2013156. <\/span><\/p>\n

D’Mello, S., Taylor, R. S., & Graesser, A. (2007). Monitoring affective trajectories during complex learning. In D. McNamara & J. Trafton (Eds.), Proceedings of the 29th Annual Cognitive Science Society <\/i>(pp. 203\u2013208). Austin, TX: Cognitive Science Society. <\/span><\/p>\n

Darwin, C. (1872). The expression of the emotions in man and animals. London: John Murray. <\/span><\/p>\n

Dillon, J., Bosch, N., Chetlur, M., Wanigasekara, N., Ambrose, G. A., Sengupta, B., & D’Mello, S. K. (2016). Student emotion, co-occurrence, and dropout in a MOOC context. In T. Barnes, M. Chi, & M. Feng (Eds.), Proceedings of the 9th International Conference on Educational Data Mining <\/i>(EDM 2016) (pp. 353\u2013357). International Educational Data Mining Society. <\/span><\/p>\n

Donnelly, P., Blanchard, N., Samei, B., Olney, A. M., Sun, X., Ward, B., . . . D’Mello, S. K. (2016a). Automatic teacher modeling from live classroom audio. In L. Aroyo, S. D’Mello, J. Vassileva, & J. Blustein (Eds.), Proceedings of the 2016 ACM International Conference on User Modeling, Adaptation, & Personalization <\/i>(ACM UMAP 2016) (pp. 45\u201353). New York: ACM. <\/span><\/p>\n

Donnelly, P., Blanchard, N., Samei, B., Olney, A. M., Sun, X., Ward, B., . . . D’Mello, S.K. (2016b). Multi-sensor modeling of teacher instructional segments in live classrooms. Proceedings of the 18th ACM International Conference on Multimodal Interaction <\/i>(ICMI 2016). New York: ACM. <\/span><\/p>\n

Ekman, P., & Friesen, W. (1978). The facial action coding system: A technique for the measurement of facial movement. Palo Alto, CA: Consulting Psychologists Press. <\/span><\/p>\n

Emotient. (2014). FACET: Facial expression recognition software. <\/span><\/p>\n

Farrington, C. A., Roderick, M., Allensworth, E., Nagaoka, J., Keyes, T. S., Johnson, D. W., & Beechum, N. O. (2012). Teaching adolescents to become learners: The role of noncognitive factors in shaping school performance: A critical literature review. Chicago, IL: University of Chicago Consortium on Chicago School Research. <\/span><\/p>\n

Forbes-Riley, K., & Litman, D. J. (2011). Benefits and challenges of real-time uncertainty detection and adaptation in a spoken dialogue computer tutor. Speech Communication, 53<\/i>(9\u201310), 1115\u20131136. doi:10.1016\/j. specom.2011.02.006<\/span><\/p>\n

Fredrickson, B., & Branigan, C. (2005). Positive emotions broaden the scope of attention and thought-action repertoires. Cognition & Emotion, 19<\/i>(3), 313\u2013332. doi:10.1080\/02699930441000238 <\/span><\/p>\n

Galla, B. M., Plummer, B. D., White, R. E., Meketon, D., D’Mello, S. K., & Duckworth, A. L. (2014). The academic diligence task (ADT): Assessing individual differences in effort on tedious but important schoolwork. Contemporary Educational Psychology, 39<\/i>(4), 314\u2013325. <\/span><\/p>\n

Graesser, A., Chipman, P., Haynes, B., & Olney, A. (2005). AutoTutor: An intelligent tutoring system with mixed-initiative dialogue. IEEE Transactions on Education, 48<\/i>(4), 612\u2013618. doi:10.1109\/TE.2005.856149 <\/span><\/p>\n

Gross, J. (2008). Emotion regulation. In M. Lewis, J. Haviland-Jones, & L. Barrett (Eds.), Handbook of emotions <\/i>(3rd ed., pp. 497\u2013512). New York: Guilford. <\/span><\/p>\n

Hess, U. (2003). Now you see it, now you don’t – the confusing case of confusion as an emotion: Commentary on Rozin and Cohen (2003). Emotion, 3<\/i>(1), 76\u201380. <\/span><\/p>\n

Isen, A. (2008). Some ways in which positive affect influences decision making and problem solving. In M. Lewis, J. Haviland-Jones, & L. Barrett (Eds.), Handbook of emotions <\/i>(3rd ed., pp. 548\u2013573). New York: Guilford. <\/span><\/p>\n

Izard, C. (2010). The many meanings\/aspects of emotion: Definitions, functions, activation, and regulation. Emotion Review, 2<\/i>(4), 363\u2013370. doi:10.1177\/1754073910374661 <\/span><\/p>\n

James, W. (1884). What is an emotion? Mind, 9<\/i>, 188\u2013205. <\/span><\/p>\n

Jayaprakash, S. M., Moody, E. W., Laur\u00eda, E. J., Regan, J. R., & Baron, J. D. (2014). Early alert of academically at-risk students: An open source analytics initiative. Journal of Learning Analytics, 1<\/i>(1), 6\u201347. <\/span><\/p>\n

Kai, S., Paquette, L., Baker, R., Bosch, N., D’Mello, S., Ocumpaugh, J., . . . Ventura, M. (2015). Comparison of face-based and interaction-based affect detectors in physics playground. In C. Romero, M. Pechenizkiy, J. Boticario, & O. Santos (Eds.), Proceedings of the 8th International Conference on Educational Data Mining <\/i>(EDM 2015) (pp. 77\u201384). International Educational Data Mining Society. <\/span><\/p>\n

Kory, J., D’Mello, S. K., & Olney, A. (2015). Motion tracker: Camera-based monitoring of bodily movements using motion silhouettes. PloS ONE, 10<\/i>(6), doi:10.1371\/journal.pone.0130293. <\/span><\/p>\n

Lench, H. C., Bench, S. W., & Flores, S. A. (2013). Searching for evidence, not a war: Reply to Lindquist, Siegel, Quigley, and Barrett (2013). Psychological Bulletin, 113<\/i>(1), 264\u2013268. <\/span><\/p>\n

Lewis, M. D. (2005). Bridging emotion theory and neurobiology through dynamic systems modeling. Behavioral and Brain Sciences, 28<\/i>(2), 169\u2013245. <\/span><\/p>\n

Lindquist, K. A., Siegel, E. H., Quigley, K. S., & Barrett, L. F. (2013). The hundred-year emotion war: Are emotions natural kinds or psychological constructions? Comment on Lench, Flores, and Bench (2011). Psychological Bulletin, 139<\/i>(1), 264\u2013268. <\/span><\/p>\n

McKeown, G. J., & Sneddon, I. (2014). Modeling continuous self-report measures of perceived emotion using generalized additive mixed models. Psychological Methods, 19<\/i>(1), 155\u2013174. <\/span><\/p>\n

Mesquita, B., & Boiger, M. (2014). Emotions in context: A sociodynamic model of emotions. Emotion Review, 6<\/i>(4), 298\u2013302. <\/span><\/p>\n

Microsoft. (2015). Kinect for Windows SDK MSDN. Retrieved 5 August 2015 from https:\/\/msdn.microsoft.com\/ en-us\/library\/dn799271.aspx <\/span><\/p>\n

Moors, A. (2014). Flavors of appraisal theories of emotion. Emotion Review, 6<\/i>(4), 303\u2013307. <\/span><\/p>\n

Nystrand, M. (1997). Opening dialogue: Understanding the dynamics of language and learning in the English classroom. Language and Literacy Series<\/i>. New York: Teachers College Press. <\/span><\/p>\n

Ocumpaugh, J., Baker, R. S., & Rodrigo, M. M. T. (2012). Baker-Rodrigo Observation Method Protocol (BROMP) 1.0. Training Manual version 1.0. New York. <\/span><\/p>\n

OECD (Organisation for Economic Co-operation and Development). (2015). PISA 2015 collaborative problem solving framework.<\/span><\/p>\n

Ohman, A., & Soares, J. (1994). Unconscious anxiety: Phobic responses to masked stimuli. Journal of Abnormal Psychology, 103<\/i>(2), 231\u2013240. <\/span><\/p>\n

Pang, B., & Lee, L. (2008). Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 2<\/i>(1\u20132), 1\u2013135. <\/span><\/p>\n

Pardos, Z., Baker, R. S. J. d., San Pedro, M. O. C. Z., & Gowda, S. M. (2013). Affective states and state tests: Investigating how affect throughout the school year predicts end of year learning outcomes. In D. Suthers, K. Verbert, E. Duval, & X. Ochoa (Eds.), Proceedings of the 3rd International Conference on Learning Analytics and Knowledge <\/i>(LAK\u201913), 8\u201312 April 2013, Leuven, Belgium (pp. 117\u2013124). New York: ACM. <\/span><\/p>\n

Parkinson, B., Fischer, A. H., & Manstead, A. S. (2004). Emotion in social relations: Cultural, group, and interpersonal processes<\/i>. Psychology Press. <\/span><\/p>\n

Pekrun, R., & Stephens, E. J. (2011). Academic emotions. In K. Harris, S. Graham, T. Urdan, S. Graham, J. Royer, & M. Zeidner (Eds.), APA educational psychology handbook, Vol 2: Individual differences and cultural and contextual factors <\/i>(pp. 3\u201331). Washington, DC: American Psychological Association. <\/span><\/p>\n

Porayska-Pomsta, K., Mavrikis, M., D’Mello, S. K., Conati, C., & Baker, R. (2013). Knowledge elicitation methods for affect modelling in education. International Journal of Artificial Intelligence in Education, 22<\/i>, 107\u2013140. <\/span><\/p>\n

Raca, M., Kidzinski, L., & Dillenbourg, P. (2015). Translating head motion into attention: Towards processing of student’s body language. Proceedings of the 8th International Conference on Educational Data Mining<\/i>. International Educational Data Mining Society. <\/span><\/p>\n

Razzaq, L., Feng, M., Nuzzo-Jones, G., Heffernan, N. T., Koedinger, K. R., Junker, B., . . . Choksey, S. (2005). The assistment project: Blending assessment and assisting. In C. Loi & G. McCalla (Eds.), Proceedings of the 12th International Conference on Artificial Intelligence in Education <\/i>(pp. 555\u2013562). Amsterdam: IOS Press. <\/span><\/p>\n

Ringeval, F., Sonderegger, A., Sauer, J., & Lalanne, D. (2013). Introducing the RECOLA multimodal corpus of remote collaborative and affective interactions. Proceedings of the 2nd International Workshop on Emotion Representation, Analysis and Synthesis in Continuous Time and Space (EmoSPACE) in conjunction with the 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition<\/i>. Washington, DC: IEEE. <\/span><\/p>\n

Rosenberg, E. (1998). Levels of analysis and the organization of affect. Review of General Psychology, 2<\/i>(3), 247\u2013270. doi:10.1037\/\/1089-2680.2.3.247 <\/span><\/p>\n

Rozin, P., & Cohen, A. (2003). High frequency of facial expressions corresponding to confusion, concentration, and worry in an analysis of maturally occurring facial expressions of Americans. Emotion, 3<\/i>, 68\u201375. <\/span><\/p>\n

Sabourin, J., Mott, B., & Lester, J. (2011). Modeling learner affect with theoretically grounded dynamic Bayesian networks. In S. D’Mello, A. Graesser, B. Schuller, & J. Martin (Eds.), Proceedings of the 4th International Conference on Affective Computing and Intelligent Interaction <\/i>(pp. 286\u2013295). Berlin\/Heidelberg: Springer-Verlag. <\/span><\/p>\n

San Pedro, M., Baker, R. S., Bowers, A. J., & Heffernan, N. T. (2013). Predicting college enrollment from student interaction with an intelligent tutoring system in middle school. In S. D’Mello, R. Calvo, & A. Olney (Eds.), Proceedings of the 6th International Conference on Educational Data Mining <\/i>(EDM 2013) (pp. 177\u2013184). International Educational Data Mining Society. <\/span><\/p>\n

Schwarz, N. (2012). Feelings-as-information theory. In P. Van Lange, A. Kruglanski, & T. Higgins (Eds.), Handbook of theories of social psychology <\/i>(pp. 289\u2013308). Thousand Oaks, CA: Sage. <\/span><\/p>\n

Shute, V. J., Ventura, M., & Kim, Y. J. (2013). Assessment and learning of qualitative physics in Newton’s playground. The Journal of Educational Research, 106<\/i>(6), 423\u2013430. <\/span><\/p>\n

Sinha, T., Jermann, P., Li, N., & Dillenbourg, P. (2014). Your click decides your fate: Inferring information processing and attrition behavior from MOOC video clickstream interactions. Paper presented at the Empirical Methods in Natural Language Processing: Workshop on Modeling Large Scale Social Interaction in Massively Open Online Courses. http:\/\/www.aclweb.org\/anthology\/W14-4102 <\/span><\/p>\n

Strain, A., & D’Mello, S. (2014). Affect regulation during learning: The enhancing effect of cognitive reappraisal. Applied Cognitive Psychology, 29<\/i>(1), 1\u201319. doi:10.1002\/acp.3049<\/span><\/p>\n

Tracy, J.L. (2014). An evolutionary approach to understanding distinct emotions. Emotion Review, 6<\/i>(4), 308\u2013 312. <\/span><\/p>\n

VanLehn, K., Siler, S., Murray, C., Yamauchi, T., & Baggett, W. (2003). Why do only some events cause learning during human tutoring? Cognition and Instruction, 21<\/i>(3), 209\u2013249. doi:10.1207\/S1532690XCI2103_01 <\/span><\/p>\n

Wang, Z., Miller, K., & Cortina, K. (2013). Using the LENA in teacher training: Promoting student involvement through automated feedback. Unterrichtswissenschaft, 4<\/i>, 290\u2013305. <\/span><\/p>\n

Wen, M., Yang, D., & Ros\u00e9, C. (2014). Sentiment analysis in MOOC discussion forums: What does it tell us? In J. Stamper, S. Pardos, M. Mavrikis, & B. M. McLaren (Eds.), Proceedings of the 7th International Conference on Educational Data Mining <\/i>(EDM 2014), 4\u20137 July, London, UK (pp. 130\u2013137): International Educational Data Mining Society. http:\/\/www.cs.cmu.edu\/~mwen\/papers\/edm2014-camera-ready.pdf <\/span><\/p>\n

Woolf, B., Burleson, W., Arroyo, I., Dragon, T., Cooper, D., & Picard, R. (2009). Affect-aware tutors: Recognizing and responding to student affect. International Journal of Learning Technology, 4<\/i>(3\/4), 129\u2013163. <\/span><\/p>\n

Yang, D., Wen, M., Howley, I., Kraut, R., & Ros\u00e9, C. (2015). Exploring the effect of confusion in discussion forums of massive open online courses. Proceedings of the 2nd ACM conference on Learning@Scale <\/i>(L@S 2015), 14\u201318 March 2015, Vancouver, BC, Canada (pp. 121\u2013130). New York: ACM. doi:10.1145\/2724660.2724677 <\/span><\/p>\n

<\/a> Zeng, Z., Pantic, M., Roisman, G., & Huang, T. (2009). A survey of affect recognition methods: Audio, visual, and spontaneous expressions. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31<\/i>(1), 39\u201358.<\/span><\/p>\n

\n
\n

1<\/a> orj. detector<\/span><\/p>\n<\/div>\n

\n

2<\/a> orj. killer app<\/span><\/p>\n<\/div>\n","protected":false},"author":1,"menu_order":6,"template":"","meta":{"pb_show_title":"on","pb_short_title":"","pb_subtitle":"","pb_authors":[],"pb_section_license":""},"chapter-type":[48],"contributor":[],"license":[],"class_list":["post-58","chapter","type-chapter","status-publish","hentry","chapter-type-numberless"],"part":46,"_links":{"self":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/chapters\/58","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/chapters"}],"about":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/wp\/v2\/types\/chapter"}],"author":[{"embeddable":true,"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/wp\/v2\/users\/1"}],"version-history":[{"count":0,"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/chapters\/58\/revisions"}],"part":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/parts\/46"}],"metadata":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/chapters\/58\/metadata\/"}],"wp:attachment":[{"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/wp\/v2\/media?parent=58"}],"wp:term":[{"taxonomy":"chapter-type","embeddable":true,"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/pressbooks\/v2\/chapter-type?post=58"},{"taxonomy":"contributor","embeddable":true,"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/wp\/v2\/contributor?post=58"},{"taxonomy":"license","embeddable":true,"href":"https:\/\/acikkitap.com.tr\/oaek\/wp-json\/wp\/v2\/license?post=58"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}