Extracting Keyphrases Using Heterogeneous Word Relations

Abstract

Extracting keyphrases from documents for providing a quick and insightful summarization is an interesting and important task, on which lots of research efforts have been laid. Most of the existing methods could be categorized as co-occurrence based, statistic-based, or semantics-based. The co-occurrence based methods do not take various word relations besides co-occurrence into full consideration. The statistic-based methods introduce more unrelated noises inevitably due to the inclusion of external text corpus, while the semantic-based methods heavily depend on the semantic meanings of words. In this paper, we propose a novel graph-based approach to extract keyphrases by considering heterogeneous latent word relations (the co-occurrence and the semantics). The underlying random walk model behind the proposed approach is made possible and reasonable by exploiting nearest neighbor documents. Extensive experiments over real data show that our method outperforms the state-of-art methods.