Near-synonyms discrimination dictionaries often overlook the nuances of near-synonyms, and lack of detailed description of the knowledge of syntactic distribution, semantic characteristics, combination restrictions and usage templates of near-synonyms. The author constructs a knowledge base of near-synonyms based on corpus, which takes collocation as the carrier and the relationship between words as the means to provide rich usage and context knowledge for near-synonyms discrimination. The proposed knowledge base contains usage information: collocation, collocational relationship, collocational frequency, distributional information, semantic features, example sentences, etc. The knowledge of syntax, semantics, and usage of near-synonyms are presented in a visual way, so it is more intuitive and more friendly to Chinese second language learners.

现有近义词辨析词典往往忽略近义词的细微用法, 缺乏对近义词的句法分布、语义特征、组合限制及使用模板的细致描述。针对该问题, 基于真实语料, 以搭配为载体, 以词与词之间的关系为手段, 构建了一个近义词辨析知识库, 为近义词提供用法和语境知识, 包括近义词的搭配词及搭配关系、搭配频率以及近义词在语料库中的句法分布信息、语义特征、语境中的例句等。该知识库用可视化方式呈现近义词的搭配词和近义词的句法、语义和用法知识, 表示更加直观, 对汉语二语学习者更友好。