Sandra Katz

Modeling the Student in Sherlock II

Student modeling—the task of building dynamic models of student ability—is fraught with uncertainty, caused by such factors as multiple sources of student errors, careless errors and lucky guesses, learning and forgetting. Various approaches have been developed in recent years to make student modeling more tractable. One approach, which is based on fuzzy set theory, aims at building imprecise, or “fuzzy” diagnostic student models; (eg. [18]). We have built upon this approach by developing techniques for representing and updating discrete student knowledge variables in our avionics troubleshooting tutor, Sherlock II. We describe these techniques and, more broadly, the student modeling component in this tutor. Future work will focus on calibrating the student modeling knowledge base and updating rules, evaluating the approach, and comparing it with other approaches to imprecise student modeling that we have implemented and are currently developing.

The cognitive skill of coaching collaboration

We observed eight experienced avionics technicians as they coached collaborating dyads who worked on problems in Sherlock 2, an intelligent tutoring system for avionics. Data analysis focused on: (1) defining the cognitive skill of coaching collaboration, (2) identifying cues that coaches use to detect and diagnose peer interaction impasses, and (3) specifying how coaches remedy these impasses. Coaching collaboration is an extension of the task of mentoring in one-on-one interactions. It involves three levels of diagnosis: diagnosing problems in the task situation, in students knowledge and problem-solving process, and in peer interactions. One-on-one instructional interactions--e.g., medical rounds consultations between an expert physician and a resident--involve only the first two levels. We describe cues that can signal to a human or automated coach that intervention is necessary and can indicate what the cause of a peer interaction impasse is. During problem solving, the avionics experts tended to focus on giving advice that would keep students on a productive solution path. They seldom addressed problems at the interaction level---e.g., by prompting a student who evidently knew what to do next to help his peer. However, experts and students used the post-practice reflective discussions to more fully address knowledge gaps, misconceptions, and interaction-level problems.

Modeling pedagogical interactions with machine learning

ZusammenfassungDieser Artikel beschreibt einen Ansatz zur Codierung und Analyse pädagogischer Interaktionen zwischen Lernenden und Experten im Bereich der Luftfahrt-Technik im Rahmen einer intelligenten tutoriellen Lernumgebung zur Fehlererkennung und -behebung (Sherlock 2). Es werden zwei prototypische maschinelle Lernverfahren dargestellt, welche bei der Analyse eines Corpus von Dialogprotokollen eingesetzt werden. Die Verfahren generieren der Diskursstruktur möglichst gut entsprechende Grammatiken. Die gefundenen Grammatiken führten zu neuen Einsichten hinsichtlich der Funktion bestimmter Sprechakte in tutoriellen Dialogen sowie zu klaren und präzisen Modellen von Erklärungsdialogen bei Diagnoseproblemen.SummaryThis paper describes our approach to coding pedagogical interactions that took place between avionics students and domain experts in an ITS for electronic troubleshooting called Sherlock 2. We also describe prototype machine learning systems that we developed to learn grammars of discourse structure. The grammars revealed instructional functions of particular speech acts that we had not been aware of, and provided concise models of explanations common in diagnostic tasks.n

KSC-PaL: a peer learning agent

We have developed an artificial agent based on a computational model of peer learning we developed That model shows that shifts in initiative are conducive to learning The peer learning agent can collaborate with a human student via dialog and actions within a graphical workspace This paper describes the architecture and implementation of the agent and the user study we conducted to evaluate the agent Results show that the agent is able to encourage shifts in initiative in order to promote learning and that students learn using the agent.