The Why2-Atlas system teaches qualitative physics by having students write paragraph-long explanations of simple mechanical phenomena.

We examine the utility of speech and lexical features for predicting student emotions in computerhuman spoken tutoring dialogues. We first annotate student turns for negative, neutral, positive and mixed emotions. We then extract acoustic-prosodic features from the speech signal, and lexical items from the transcribed or recognized speech. We compare the results of machine learning experiments using these features alone or in combination to predict various categorizations of the annotated student emotions. Our best results yield a 19-36 % relative improvement in error reduction over a baseline. Finally, we compare our results with emotion prediction in human-human tutoring dialogues. 1

This project augments an existing intelligent tutoring system (AutoTutor) that helps learners construct explanations by interacting with them in natural language and helping them use simulation environments. The research aims to develop an agile learning environment that is sensitive to a learner’s affective state, presuming that this will promote learning. We integrate state-of-the-art, nonintrusive, affect-sensing technology with AutoTutor in an endeavor to classify emotions on the bases of facial expressions, gross body movements, and conversational cues. This paper sketches our broad theoretical approach, our methods for data collection and evaluation, and our emotion classification techniques.

Many interactive learning environments (ILEs) offer on-demand help, intended to positively influence learning. Recent studies report evidence that although effective help-seeking behavior in ILEs is related to better learning outcomes, learners are not using help facilities effectively. This selective review (a) examines theoretical perspectives on the role of on-demand help in ILEs, (b) reviews literature on the relations between help seeking and learning in ILEs, and (c) identifies reasons for the lack of effective help use. We review the effect of system-related factors, of student-related factors, and of interactions between these factors. The interaction between metacognitive skills and cognitive factors is important for appropriate help seeking, as are a potentially large space of system-related factors as well as interactions among learner- and system-related factors. We suggest directions for future research.

Click here to download paper in PDF format This article presents a vision of the near future in which computer interaction is characterized by natural face-toface conversations with lifelike characters that speak, emote and gesture. These animated agents will converse with people much like people converse effectively with assistants in a variety of focused applications. Despite the research advances required to realize this vision, and the lack of strong experimental evidence that animated agents improve human computer interaction, we argue that initial prototypes of perceptive animated interfaces can be developed today, and that the resulting systems will provide more effective and engaging communication experiences than existing systems. In support of this hypothesis, we first describe initial experiments using an animated character to teach speech and language skills to children with hearing problems, and classroom subject and social skills to children with autistic spectrum disorder. We then show how existing dialogue system architectures can be transformed into perceptive animated interfaces by integrating computer vision and animation capabilities. We conclude by describing the Colorado Literacy Tutor, a computer-based literacy program that provides an ideal test bed for research and development of perceptive animated interfaces, and consider next steps required to realize the vision.

systems Previous studies have shown that self-explanation is an effective metacognitive strategy and can be supported effectively by intelligent tutoring systems. It is plausible however that students may learn even more effectively when stating explanations in their own words and when receiving tutoring focused on their explanations. We are developing the Geometry Explanation Tutor in order to test this hypothesis. This system requires that students provide general explanations of problem-solving steps in their own words. It helps them through a restricted form of dialogue to improve explanations and arrive at explanations that are mathematically precise. Based on data collected during a pilot study in which the tutor was used for two class periods in a junior high school, the techniques we have chosen to implement the dialogue system, namely a knowledge-based approach to natural language understanding and classification of student explanation, seem up to the task. There are a number of ways in which the system could be improved within the current architecture.

Some of the recent electronic learning environments have moved beyond the conventional delivery of text, multimedia, and objective tests. There are systems with animated conversational agents, intelligent adaptive tutoring, interactive simulations, and other features designed to engage learners and promote deeper comprehension. One system is AutoTutor, a learning environment that tutors students by holding a conversation in natural language. AutoTutor’s design was inspired by explanation-based constructivist theories of learning, intelligent tutoring systems that adaptively respond to student knowledge, and empirical research on dialogue patterns in tutorial discourse. AutoTutor presents challenging questions and then engages in mixed initiative dialogue that guides the student in building an answer. It provides feedback to the student on what the student types in (positive, neutral, negative feedback), pumps the student for more information, prompts the student to fill in missing words, gives hints, fills in missing information, identifies and

Abstract. Previous research has shown that self-explanation can be supported effectively in an intelligent tutoring system by simple means such as menus. We now focus on the hypothesis that natural language dialogue is an even more effective way to support self-explanation. We have developed the Geometry Explanation Tutor, which helps students to state explanations of their problem-solving steps in their own words. In a classroom study involving 71 advanced students, we found that students who explained problem-solving steps in a dialogue with the tutor did not learn better overall than students who explained by means of a menu, but did learn better to state explanations. Second, examining a subset of 700 student explanations, students who received higher-quality feedback from the system made greater progress in their dialogues and learned more, providing some measure of confidence that progress is a useful intermediate variable to guide further system development. Finally, students who tended to reference specific problem elements in their explanations, rather than state a general problemsolving principle, had lower learning gains than other students. Such explanations may be indicative of an earlier developmental level. 1

We added a sentence planning component to an existing ITS that teaches students how to troubleshoot mechanical systems.

Teaching and learning are highly social activities. Seminal psychologists such as Vygotsky, Piaget, and Bandura have theorized that social interaction is a key mechanism in the process of learning and development. In particular, the benefits of peer interaction for learning and motivation in classrooms have been broadly demonstrated through empirical studies. Hence, it would be valuable if computer-based environments could support a mechanism for a peer-interaction. Though no claim of peer equivalence is made, pedagogical agents as learning companions (PALs)-- animated digital characters functioning to simulate human-peer-like interaction-- might provide an opportunity to simulate such social interaction in computer-based learning. The purpose of this paper is first to ground the instructional potential of PALs in several social-cognitive theories, which include distributed cognition, social interaction, and Bandura’s social-cognitive theory. The paper discusses how specific concepts of the theories might support various instructional functions of PALs, acknowledging concepts that PALs cannot address. Next, based on the theoretical perspectives, the paper suggests seven key constituents for designing PALs that in human-peer interactions have proven significant: PAL competency, interaction type, gender, affect, ethnicity, multiplicity, and feedback. Finally, the paper reviews the current status of PAL research with respect to these constituents and suggests where further empirical research is necessary.