Prerequisite relations describe a fundamental relation among concepts in education and are useful for representing concept graphs for various knowledge disciplines. My research focuses on developing machine learning methods for discovering concept prerequisite relations. Different from existing methods in educational data mining which are largely based on student assessment data, our approaches make use of various educational data sources such as Wikipedia, textbooks, and course prerequisites. The goal is to have a scalable learning model for measuring general prerequisite relations among concepts.

• Please check out our recent work on active learning for concept prerequisite learning (aaai'18; CoRR) to learn more!
• Here is a paper list on concept prerequisite learning, selected from AI/NLP conferences.

Related Publications:

• [EDM'19] Active Learning of Strict Partial Orders: A Case Study on Concept Prerequisite Relations
• [CoRR] Active Learning of Strict Partial Orders: A Case Study on Concept Prerequisite Relations
• [AAAI'18] Investigating Active Learning for Concept Prerequisite Learning
• [EMNLP'15] Measuring Prerequisite Relations Among Concepts.
• [AAAI'17] Recovering Concept Prerequisite Relations from University Course Dependencies
• [CIKM'16] Using Prerequisites to Extract Concept Maps from Textbooks.

Collaborating with the TLT department at Penn State, we apply information retrieval methods to dynamically generate "zero-cost" materials to support learning. The end goal is to have a system that can search various OERs, based on a set of user-generated criteria, and return various resources that can be combined, remixed, and re-used to support specific learning goals. The current system is designed to search wikipedia, and return results to users based on the search criteria. The more times users iterate the search (accept or reject results), the better the search becomes. As of December 2016, BBookX has 416 users, that created 419 books. Over 47000 searches were executed within BBookX, resulting in 1086 book chapters. Please check out this link to learn more.
Media Coverage: Tech Times, Tech Republic, Philadelphia Inquirer, Campus Technology, Center for Digital Education, Education World, Penn State News, The Daily Collegian, T.H.E. Journal, EdSurge, etc.
Presentations:

• CLIEDE 2017, Sep 2017, University Park, PA, USA
• WWW 2016, April 2016, Montreal, Canada
• AAAI 2016, Feb 2016, Phoenix, AZ, USA
• Penn State EdTech Network Summit, Nov 2015, University Park, PA, USA
• DocEng 2015, Sep 2015, Lausanne, Switzerland

Related Publications:

• [AAAI'16] BBookX: Building Online Open Books for Personalized Learning.
• [WebED'16'] BBookX: Design of an Automated Web-based Recommender System for the Creation of Open Learning Content.
• [DocEng'15] BBookX: An Automatic Book Creation Framework.
• [DocEng'15] Concept Hierarchy Extraction from Textbooks.