BEGIN:VCALENDAR
BEGIN:VEVENT
DTSTART:20101126T200000Z
DTEND:20101126T210000Z
SUMMARY:ICS Colloquium - Dr. Kasia Muldner
DESCRIPTION:Scaffolding Learning with Adaptive Technologies: Challenges and Directions



Dr. Kasia Muldner from Arizona State University



Intelligent Tutoring Systems (ITSs) are computer applications that employ artificial intelligence techniques to instruct students in an “intelligent way”. Although there isn’t an accepted definition of the term “intelligent”, a characteristic shared by many ITS is that they possess knowledge and reasoning capabilities to adapt the instruction to the needs of each individual student. This functionality is motivated by Bloom’s seminal research demonstrating that when students receive tailored one-on-one instruction, they perform two standard deviations over standard classroom instruction. Although ITSs have made significant progress in the last decade, many challenges remain. Some of these challenges include: (1) how to support the full spectrum of student needs, spanning not only the cognitive, but also meta-cognitive and affective, (2) how to unobtrusively and seamlessly assess students during their interaction with an ITS, and (3) how to leverage and extend as needed theories from cognitive science and psychology to realize the design of these complex systems.



In this talk, I will describe preliminary steps in addressing some of these challenges, including how the integration of cognitive science, artificial intelligence and human-computer interaction can guide the design and evaluation of ITSs. The first ITS I will describe corresponds to a tutor that supports meta-cognitive skills needed to learn from Analogical Problem Solving (APS) in the target domain of Newtonian physics. The design of this tutor is based on cognitive theories of learning from examples, including Anderson’s ACT-R and VanLehn’s Cascade; the implementation relies on probabilistic modeling and decision-theoretic methods in order to optimize instruction. A second set of work I will describe focuses on the affective dimension, namely how various sensing devices can be used to collect information on students’ affect in order to tailor affective interventions to students’ needs. I will also present recent results from educational data mining based on several years of students' interaction with one particular ITS; our findings highlight that students, rather than lesson features are key predictors of un-constructive student behaviors.
LOCATION:
URL:
END:VEVENT
END:VCALENDAR