Disambiguation for Chinese network encyclopedias based on semi-topic model

Network encyclopedias have been a major method for searching information or knowledge in our daily life. There are many hyperlinks pointing other entries to make readers better comprehend in network encyclopedias, but these may be ambiguous. Some algorithms have been proposed for disambiguation in natural language processing. Most of those are based on English, so in this paper we propose a novel algorithm based on Chinese. The paper focuses on the semantics and topic model to analyze the relationship between entries and the experiment indicates that it is more effective than traditional algorithm.