Abstract—Due to popularization of SNS and increase of use of WEB, people have to deal with large number of text data. However, it is difficult to process huge text data manually. For this problem, the classification methods based on machine learning is considered to be applicable. As a method of document classification, WEBSOM and its variations can visualize the relations among the documents as the similar documents are classified closely on the 2 dimension plane, and they will present good usability to the user because of their visualization ability. In this paper, the document classification method based on SOM and Word2Vec model, which can reduce the computational costs as to be executed on personal computers and can enhance the visualization ability. The performance of the proposed method is examined in the experiments using general collection of documents, and DNA sequences as the sample data.
Index Terms—Self Organizing Map (SOM), Word2Vec, documents classification.
The authors are with the Graduate School of Science and Engineering Saga University, 1 Honjyo Saga, 840-8502, Japan (e-mail: 17578035@edu.cc.saga-u.ac.jp, hiro@dna.ec.saga-u.ac.jp).
Cite: Koki Yoshioka and Hiroshi Dozono, "The Classification of the Documents Based on Word2Vec and 2-Layer Self Organizing Maps," International Journal of Machine Learning and Computing vol. 8, no. 3, pp. 252-255, 2018.