AIED 2017 AutoTutor Tutorial
This tutorial focuses on the authoring process of AutoTutor lessons, including AutoTutor dialogues and trialogues, conversation elements, media elements, conversation rules, and template-based authoring. Participants need to bring Windows laptops. A Windows authoring tool will be released on site. An example AutoTutor lesson will be provided to participants. Each participant will create their own AutoTutor lesson by modifying the example lesson.
Decades of research and development on intelligent tutoring systems (ITS) have resulted in a flourishing field within educational technology. Many tutoring systems provide rich media content that students can interact with, including multiple choice answer selection, drag and drop objects, rearranging objects, assembling objects, and so on.
Additionally, a small subset of ITSs have attempted to interact with students using conversational dialogue. Delivering content with conversation is always attractive to content authors and students, and research has shown that delivering content through conversation is much more effective than through reading text alone.
Difficulties with Natural Language Processing
Unfortunately, creating conversational content is difficult for multiple reasons:
- In order to have a natural language conversation with a student, the ITS has to be able to “understand” the student’s natural language input. There isn’t a perfect natural language algorithm that can truly understand users' free language.
- Preparing tutoring speeches for conversations is effortful; an author needs to consider many (if not infinite) responses to all possible student inputs.
- It is hard to create and test conversation rules. Conversation rules decide when a prepared utterance is spoken. Since the tutoring conversations are often combined with other displayed content, such as text, image, video, etc., conversation rules need to take into account everything that has taken place in the learning environment, in addition to students' natural language inputs.
- Other difficulties involve talking head features and
behaviors, including speech synthesizing, lip synchronization, emotion,
gesture, speech recognition, emotion detection, and so on.
Since the 1990s, the AutoTutor team at the Institute for Intelligent systems (IIS) at the University of Memphis has been providing solutions to these problems.
AutoTutor helps students learn by holding deep reasoning conversations. An AutoTutor conversation often starts with a main question about a certain topic. The goal of the conversation is to help the student construct an acceptable answer to the main question. Rather than simply telling the student the answer, AutoTutor asks a sequence of questions (hints, prompts) that target specific concepts involved in the ideal answer to the main question. AutoTutor systems respond to students' natural language input, as well as other interactions, such as making a choice, arranging some objects in the learning environment, etc.
AutoTutor in Action
AutoTutor has been incorporated into over a dozen conversational ITSs, including:
- computer literacy tutor
- conceptual physics tutor
- critical thinking tutor (OperationARIES!)
- adult literacy tutor (CSAL)
- electronics tutor (ElectornixTutor)
- and more!
Additionally, a team at National Taichung University of Education has developed a Chinese language tutor.
If you are interested in learning more about AutoTutor, join us at the AIED 2017 AT Workshop!
Planned Schedule (Not final)
Session 1: Introduction to AutoTutor.
9:00-9:15 Introduction – Introdcution of presenters and participants
9:15-10:30 Overview and Demo of AutoTutor Systems
Session 2: AutoTutor Script Authoring Tool.
11:00-12:30 A step by step guidance to creating an AutoTutor lesson
Session 3: Team practicing.
14:00-15:30 Each teams (2-3 people) produce an AutoTutor lesson by modifying provided example.
Session 4: Team report and discussions.
16:00-17:00 Team report
Graesser, A.C. (2016). Conversations with AutoTutor help students learn. International Journal of Artificial Intelligence in Education, 26, 124-132.
Nye, B.D., Graesser, A.C., & Hu, X. (2014). AutoTutor and family: A review of 17 years of natural language tutoring. International Journal of Artificial Intelligence in Education, 24(4), 427-469. doi:10.1007/s40593-014-0029-5
Graesser, A. C., Li, H., & Forsyth, C. (2014). Learning by communicating in natural language with conversational agents. Current Directions in Psychological Science, 23(5), 374-380. doi:
Graesser, A. C., D'Mello, S. K., & Strain, A. (2011). Computer agents that help students learn with intelligent strategies and emotional sensitivity. Philippine Computing Journal, 6(2), 1-8.
D'Mello, S. K., & Graesser, A. C. (2010). Multimodal semi-automated affect detection from conversational cues, gross body language, and facial features. User Modeling and User-Adapted Interaction, 20(2), 147-187.
Graesser, A. C., Jackson, G. T., & McDaniel, B. (2007). AutoTutor holds conversations with learners that are responsive to their cognitive and emotional states. Educational Technology, 47, 19-22.
VanLehn, K., Graesser, A. C., Jackson, G. T., Jordan, P., Olney, A. M., & Rose, C. (2007). When are tutorial dialogues more effective than reading? Cognitive Science, 31, 3-62.
- Zhiqiang Cai, Research Assistant Professor, The University of Memphis, USA
- Xiangen Hu, Professor, The University of Memphis, USA; Normal University of Central China, China
- Keith Shubeck, Ph.D. Candidate, The University of Memphis, USA
- Kai-chih Bai, Ph.D. Candidate, National Taichung University of Education, Taiwan
- Art Graesser, Professor, The University of Memphis, USA
- Bor-Chen Kuo, Professor, National Taichung University of Education, Taiwan
- Chen-Huei Liao, Professor, National Taichung University of Education, Taiwan