This Major Qualifying Project specialized in analyzing textual feedback messages for problems on the ASSISTments platform. Initially, work focused on designing and implementing a randomized control trial to measure the effectiveness of common wrong answer feedback messages. With work finished on the design of the RCT, focus shifted to generating a set of features using natural language processing techniques for each CWA feedback message, to be used after the completion of the RCT. After the generation of these features, the final stages of this project applied a similar method to hint messages in ASSISTments’ TeacherASSIST system. NLP generated features were generated in order to analyze the effectiveness of hint messages, measured using next problem correctness as the dependent measure.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Lee, Morgan
• Babalola, Benjamin

Publisher

• Worcester Polytechnic Institute

Identifier

• E-project-032322-200119
• 53286

Mot-clé

• Machine Learning
• Natural Language Processing
• Education
• ASSISTments

Advisor

• Heffernan, Neil

Year

• 2022

Sponsor

• ASSISTments

Date created

• 2022-03-23

Resource type

• Major Qualifying Project

Major

• Computer Science

Rights statement

• In Copyright

License

• Creative Commons BY Attribution 4.0 International