This paper describes our recent attempts to incorporate human-like conversational behaviors into AutoTutor, an animated pedagogical agent that simulates the dialog moves of human tutors. The first section of the paper provides a brief overview of AutoTutor. The second section describes a set of conversational behaviors that are being incorporated into AutoTutor. In particular, behaviors that involve head movements, facial expressions, intonation variation, gaze behaviors, and blinking patterns are discussed. What is AutoTutor? AutoTutor is an animated pedagogical agent that engages in a conversation with the learner while simulating the dialog moves of human tutors. AutoTutor is currently designed to help college students learn about topics that are typically covered in an introductory computer literacy course (e.g., hardware, operating systems, the Internet). AutoTutor's architecture is comprised of five major modules: (1) an animated agent, (2) a curriculum script, (3) lang...

Users of educational hypertext are faced with the challenge of creating meaning both within and between texts. Cohesion is an important factor contributing to whether a reader is able to capture meaning and comprehend text. When readers are required to fill in conceptual gaps in text, comprehension can fail if they do not have sufficient knowledge. Cohesion helps low-knowledge readers to create a more coherent mental representation of the text. However, text that is too cohesive can inhibit active processing, and thus reduce coherence for more knowledgeable readers. Similar patterns have been found for hypertext, which requires readers to create coherence between multiple electronic texts. Domain novices are in greater need of explicit pointers to important links between documents and gain from having less control over system navigation. Domain experts are in less need of scaffolding within the system. We discuss the use of a multimedia reading strategy training program to help low-knowledge readers better understand less cohesive text. Finally, we discuss four principles to guide hypertext development geared toward improving coherence and metacognitive engagement.

The Tutoring Research Group at the University of Memphis has developed a computer tutor (called AutoTutor) that simulates the discourse patterns and pedagogical strategies of a typical human tutor. The dialog tactics were based on a previous project that dissected 100 hours of naturalistic tutoring sessions. AutoTutor is currently targeted for college students in introductory computer literacy courses, who learn the fundamentals of hardware, operating systems, and the Internet. Instead of merely being an information delivery system, AutoTutor serves as a discourse prosthesis (or collaborative scaffold) that assists the student in actively constructing knowledge. A dialog manager coordinates the conversation that occurs between a learner and a pedagogical agent, whereas lesson content and world knowledge are represented in a curriculum script and latent semantic analysis. The agent is a talking head with discoursesensitive facial expressions and synthesized speech. Evaluations of AutoTutor have shown that the tutoring system improves learning and memory of the lessons by .5 to .6 standard deviation units. This article describes the components of AutoTutor and contrasts two versions that follow somewhat different teaching tactics.