JIITA, vol.6 no.2, p.567-573, 2022, DOI:
Sang C. Suh, James A. Avery
Department of Computer Science, Texas A&M University–Commerce, U.S.A.
Abstract: Transfer credit evaluation is the process of mapping postsecondary courses taken by a student at one institution to courses at another institution to which they have transferred. It is a time-consuming and potentially errorprone process. While best practices and specialized information systems exist to help facilitate the process, there is still room for improvement in determining equivalencies for courses that are previously unencountered or with outdated equivalencies. This study evaluates three NLP techniques that could assist with the evaluation of transfer courses by ranking courses based on the semantic similarity of their course descriptions. It was found that cosine similarity between Word2Vec embeddings can be used to rank course descriptions accurately and quickly enough to be useful in a real-world situation.
Keywords: Natural Language Processing, Transfer Credit Evaluation, Word2Vec, Text Embeddings, Semantic Similarity, WordNet
Fullpaper: