Finding the Semantic Relationship Between Wikipedia Articles Based on a Useful Entry Relationship

Wikipedia is the largest online Internet encyclopedia, and everyone can create and edit different articles. On the one hand, because it contains huge amounts of articles and there are many different language versions, it often faces synonymous and polysemy problems. On the other hand, since some of the similar Wikipedia articles may have the same topic of discussion, it needs a suitable way to identify effectively the semantic relationships between articles. This paper first uses three well-known semantic analysis models LSA, PLSA, and LDA as evaluation benchmarks. Then, it uses the entry relationship between Wikipedia articles to design its model. According to the experimental results and analysis, its model has high performance and low cost characteristics compared with other models. The advantages of its model are as follows: (1) it is a good model for finding the semantic relationships between Wikipedia articles; (2) it is suitable for dealing with huge amounts of documentation.