Study on quantifying the semantic correlative relation of Chinese word-pair
Maosheng Zhong,Hui Liu,Lei Liu
2009-01-01
Journal of Information and Computational Science
Abstract:The quantified research of the semantic relation between words is an important and basic subtask for some natural language processing, such as Text Classification, Information Retrieval and Statistical-based Language Models. Generally, there are three types of semantic relations between words, namely, Synonymy Relation, Hyponymy Relation and correlative Relation. The existing quantified research into semantic relations between words mostly focuses on how to quantify the Synonymy Relation. We studied on quantifying Semantic Correlative Relation of word-pair in this paper, and analyzed the drawbacks of such approaches as Probability Covariance, Co-occurrence Frequency and Mutual Information, for the task of quantifying Semantic Correlative Relation between words. Then, we presented a novel approach to quantify the Semantic Correlative Relation of word-pair, and constructed Bipartite Graph of Lexical Correlative Relation for measuring the indirectly correlative relation between words. Moreover, the approach can solve the problem of how to quantify the Semantic Correlative Relation of word-pair, which do not co-occur in the Corpus. The results showed that our approach is more feasible than Mutual Information means. At the same time, the approach can obtain a result of relatively reasonable tendency on quantified semantic correlative relation between a specific word and the other words. Copyright ©2009 Binary Information Press.