NC State University

Department of Computer Science Colloquia Series 1998-99

Date: Friday, April 16, 1999
Time: 3: 30 PM (talk)
Place: Withers 402A, NCSU Historical Campus (click for courtesy parking request)

Speaker: Nancy Green, School of Computer Science, CMU

Interactive Multimedia Argumentation in AutoBrief, an Intelligent Multimedia Presentation System

Abstract: AutoBrief is an intelligent interactive presentation system being developed by researchers at Carnegie Mellon University and the University of Pittsburgh. This talk will focus on how AutoBrief automatically generates interactive arguments expressed in a combination of text and information graphics (non-pictorial graphics such as bar charts) from abstract and quantitative data. The current prototype version of AutoBrief generates presentations about transportation schedules produced by a scheduling program. Typically, users must run the scheduling program over and over, trying different combinations of requirements and resources, until a satisfactory schedule is produced.

The goal of the presentation system is to reduce the burden on the user in this process by providing summaries, pointing out possible problems, and suggesting possible changes to make. AutoBrief must be able to produce arguments justifying its recommendations, based on the data that led to the recommendation. An argument is expressed using both text and graphics to take advantage of the strengths of each medium. In addition to providing information, the presentation is highly interactive, enabling the user to use presentation elements (both text and graphics) from AutoBrief to control other applications in the graphical user-interface environment.

Short Bio: Nancy Green is a Systems Scientist in the Human-Computer Interaction Institute at Carnegie Mellon University. Her postdoctoral research in the Computer Science Department at Carnegie Mellon University was on using a computational cognitive architecture (SOAR) to create a unified approach to discourse planning/learning and plan recognition for a dialogue interface to an intelligent agent. For her dissertation research at the Department of Computer and Information Sciences at the University of Delaware, she developed a computational model for generating and interpreting indirect answers, a use of inference occurring frequently in human-human dialogue. She also holds a Master of Arts in Linguistics from the University of North Carolina at Chapel Hill and a Bachelor of Arts from North Carolina State University. Her current research interests are in human-computer interaction and natural language processing, including the use of spoken and written language in multimedia and multimodal interfaces, and argumentation in computer media.

Colloquia Home Page.