Event Title

A Dialog System to Supplement Student Advising

Presenter Information

Brian McMahan, Minnesota State University, MankatoFollow

Location

CSU 284A

Start Date

5-4-2011 9:00 AM

End Date

5-4-2011 10:30 AM

Student's Major

Computer Information Science

Student's College

Science, Engineering and Technology

Mentor's Name

Rebecca Bates

Mentor's Department

Integrated Engineering

Mentor's College

Science, Engineering and Technology

Description

Dialog systems are an implementation of natural language processing theory which allows use of natural sentences for communication with a computer system. The purpose of this project was to design and implement a dialog system to augment university student advising, which is a relatively narrow domain of possible questions and responses. To further narrow the domain the dialog system focused on prescriptive advising rather than developmental advising. To better understand the domain, a professional advisor was recorded during a mock advising session to model student-advisor interaction. The natural phrases from the mock advising session were transcribed. The session was analyzed to develop the conversation sub-goals to gain data for organizing dialog. Phrases were encoded using Artificial Intelligence Markup Language (AIML). The system included a state manger to facilitate the conversation and AIML interaction. The purpose of the state manager is to control the flow of the conversation and ensure that the sub-goals are satisfied. A rule-based expert system was implemented to process the conversation results and make advising decisions, generating sentences and a visual representation of a proposed schedule to return to the student.

The system was evaluated by an advisor for correctness and tested on several students for feedback on ease of use and likelihood of using the system. Future work will include expanding the conversational ability and the knowledge domain so that the system can be useful at other universities.