Research Area:  Machine Learning
Recent methods for learning vector space representations of words have succeeded in capturing fine-grained semantic and syntactic regularities using vector arithmetic, but the origin of these regularities has remained opaque. We analyze and make explicit the model properties needed for such regularities to emerge in word vectors. The result is a new global logbilinear regression model that combines the advantages of the two major model families in the literature: global matrix factorization and local context window methods. Our model efficiently leverages statistical information by training only on the nonzero elements in a word-word co-occurrence matrix, rather than on the entire sparse matrix or on individual context windows in a large corpus. The model produces a vector space with meaningful substructure, as evidenced by its performance of 75% on a recent word analogy task. It also outperforms related models on similarity tasks and named entity recognition.
Keywords:  
Author(s) Name:   Jeffrey Pennington, Richard Socher, Christopher D. Manning
Journal name:  
Conferrence name:  Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing
Publisher name:   ACL Anthology
DOI:  10.3115/v1/D14-1162
Volume Information:  pp. 1532-1543
Paper Link:   https://aclanthology.org/D14-1162/