ERIC Number: ED237054
Record Type: Non-Journal
Publication Date: 1982-Dec
Reference Count: N/A
MENO-II: An AI-Based Programming Tutor.
Soloway, Elliot; And Others
This report examines the features and performance of the BUG-FINDing component of MENO-II, a computer-based tutor for beginning PASCAL programming students. A discussion of the use of artificial intelligence techniques is followed by a summary of the system status and objectives. The two main components of MENO-II are described, beginning with the BUG-FINDer, which draws on a database of 18 common bug or error types, represented as templates, and attempts to match these templates against its analysis of a student's program. The TUTOR component then infers the student's misconception from the type of error identified. Examples of TUTOR's analyses and interactions are presented. A preliminary evaluation of the BUG-FINDing component used the first syntactically-correct, looping program produced by 20 students randomly selected from 116 volunteers in an introductory PASCAL programming class. Results indicate that the BUG-FINDER found only 22% of the errors and was correct 55% of the time in its analysis. Examples of both correct and incorrect analyses are examined, and reasons for the poor performance of the system are suggested. Plans to extend the MENO-II are discussed, and 25 references are listed. (LMM)
Publication Type: Reports - Descriptive
Education Level: N/A
Sponsor: Office of Naval Research, Arlington, VA. Personnel and Training Research Programs Office.
Authoring Institution: Yale Univ., New Haven, CT. Dept. of Computer Science.