Chinese open information extraction based on DBMCSS in the field of national information resources

Abstract

Abstract Binary entity relationship tuples can be applied in many fields such as knowledge base construction, data mining, pattern extraction, and so on. The purpose of entity relationship mining is discovering and identifying the semantic relationship. As the relationship between entities are different from the general domain, using supervise learning methods to extract entity relationships in the field of ethnicity is difficult. After research, we find that some words can be used in the context of a sentence to describe the semantic relationship. In order to salve the existing difficulties of building tagged corpus and the predefined entities-relationships model, this paper proposes a method of density-based multi-clustering clustering of semantic similarity (DBMCSS) to mine the binary entity relationship tuples from the Chinese national information corpus, which can extract entity relationships without a training corpus.