Improving semantic similarity retrieval with word embeddings

摘要

Word similarity matchmaking is one of the core research areas of information retrieval. The existingmethods based on a synonym dictionary would lead to the problem of semantic gap, whichcould be caused by the absence of synonyms. To address this problem, we improve semantic similarityretrieval by incorporating word embeddings. Especially, word embeddings are trained byWord2Vec and then use them to depict the semantic similarity between words. Experiments areconducted on two different datasets, ie, one is a public long text dataset (ie, Reuters-21578), andthe other is a short text dataset (ie, 120ask) collected from a healthcare community. The experimentalresults on the twodatasets showthat the proposed method further improves the accuracyof the similarity retrieval.