Armand Joulin and Tomas Mikolov ▪ Enriching Word Vectors with Subword Information ▪ Proceedings of TACL 2017, Transactions of the Association for Computational Linguistics ▪ pp.135–146 ▪ キーワード ▪ WordEmbedding, fasttext 2
Sampling : 5 ▪ Character n-gram : 3-gram から 6-gramまで使用 ▪ 実験 ▪ Human similarity judgement ▪ Word analogy tasks ▪ Comparison with morphological representations ▪ Effect of the size of the training data ▪ Effect of the size of n-grams