基于多种方法相融合的词义消歧的研究

Word Sense Disambiguation Research Based on Fusion of Varieties of Methods

词义消歧是自然语言处理领域的一个关键课题,对众多的研究领域都具有十分重要的理论意义和实践意义。通过对现有的传统的词义消歧方法的分析研究,发现传统的词义消歧方法都存在一个共同的缺点,数据的稀疏问题比较严重。该文提出一种基于多种方法相融合的词义消歧方法,通过实验证实该方法不仅在词义消歧的正确率高于一般传统的词义消歧方法,同时在一定程度上缓解了数据稀疏问题。

Word sense disambiguation(WSD) research is a crucial problem in the field of natural language procession,that it has very important theoretical and practical significance among many researching areas.The result shows that the traditional wsd methods all have a common drawback,through researching on the existing traditional methods of WSD,and the shortcoming of data sparse is great.This paper proposes a method of WSD research based on fusion of varieties of methods.This method improved the precision of WSD compaired with the traditional technique,according to the experiment.Of course,the method suggested in this paper solved the proplem of data sparse in some level.