Host: The Japanese Society for Cognitive Psychology
The NTT Kanji database (Aman and Kondo, 1999) is one of the most popular kanji datasets. However, it has been long ever since it published. The word2vec (Mikolov, et al., 2013) was proposed based on large vocabulary dataset. In spite of their popularity, any comparisons have not been tried so far. We tried to figure out differences between them in terms of several ways.