Using Word2vec to Match Knowledge Points and Test Questions: A Case Study
Pei-Yan Yuan, A-Mei Du, Chuan Xi Wang · 2020
The development of artificial intelligence and machine learning has strengthened the deep integration of information technology and education. Test questions are important for students' daily practice as online learning resources. Therefore, studying the matching method between knowledge points and test questions has great meaning both in theory and reality. In this paper, the cosine distance is used to calculate the similarity between the knowledge points text and the test questions text. The Word2vec model is used to distill the character words to obtain word vectors. The similarity in vector space is used to represent the semantic similarity of text. On the similarity computing, we compared the method based on Word2vec with TF-IDF word frequency statistics. It is verified that the method based on Word2vec can quickly find the relationship between knowledge points and test questions, so as to provide technical support for education and teaching.