Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD) Regional Conference
23 May 2017
Entity set expansion (ESE) is the problem that expands asmall set of seed entities into a more complete set, entities of which havecommon traits. As a popular data mining task, ESE has been widelyused in many applications, such as dictionary construction and querysuggestion. Contemporary ESE mainly utilizes text and Web informa-tion. That is, the intrinsic relation among entities is inferred from theiroccurrences in text or Web. With the surge of knowledge graph in recentyears, it is possible to extend entities according to their occurrences inknowledge graph. In this paper, we consider the knowledge graph as aheterogeneous information network (HIN) that contains dierent typesof objects and links, and propose a novel method, called MP ESE, toextend entities in the HIN. The MP ESE employs meta paths, a relationsequence connecting entities, in HIN to capture the implicit commontraits of seed entities, and an automatic meta path generation method,called SMPG, is provided to exploit the potential relations among enti-ties. With these generated and weighted meta paths, the MP ESE caneectively extend entities. Experiments on real datasets validate the ef-fectiveness of MP ESE.