Embedded Lexica: Extracting Topical Dictionaries from Unlabeled Corpora using Word Embeddings

• In Event: Machine Learning Methods for Text

Brief Overview

In this paper, I present a novel algorithm, conclust, that applies word embeddings towards extracting topically-related dictionaries from unlabeled text using a small number of user-provided seed words and fitted word embeddings models.

Author

• Patrick J Chester, University of California-San Diego