AIED 2015 - Notes on Affect
26 Jun 2015The two themes that I payed closed attention to at the AIED 2015 conference in Madrid were i) Affect, Motivation, and Engagement, and ii) Natural Language Approaches.
Affect
Most of the work was on measuring the impact of affective states on student learning. Some notable papers were:
-
Carol Forsyth (ETS) and Art Graesser’s work using AutoTutor to evaluate the impact of tutor-agent’s and teacher-agent’s mood on student learning. This was interesting in the operationalization of mood, which is longer than an emotion, and can also be referred to as an affective state. This is measured along the dimensions of Arousal and Valence.
-
Blair Lehman (ETS) and Art Graesser’s work on the mechanics of how students resolve confusion, and the conditions in which it can impact learning (also using AutoTutor)
-
Papousek and Pelanek’s (University of Brno) work on measuring the impact on motivation in their adaptive web-based geography tutor. Best short paper award
-
Harley, Carter and Azevedo’s work on adapting the emotions of MetaTutor agents based on personality profile of the students
-
This paper was different as it was really about detecting affective states. Bosch, D’Mello, Ocumpaugh, Shute’s work on the temporal generalizability of of face based affect detection in Noisy classroom environments. This is good example of how the BROMP classroom observation protocol is combined with data mining techniques. Best Paper award.
-
The winner of the Interactive Event (demo session) was a big Europeean project called iTalk2Learn, which I can only summarize as PhET, with a microphone/earphones attached. Students engaged with exploratory learning environments while getting adaptive feedback based on their affective state. Best demo award.