论文部分内容阅读
[目的 /意义]针对共词分析存在的普遍问题,提出一种基于细粒度语义分析的共词网络构建与分析方法。[方法 /过程]借助SemRep实现源文本主题概念及其语义关系的规范化抽取并由此构建语义共词网络,然后以节点的中心度和边的频次为指标对内容特征词进行抽取,利用UMLS语义网络规定的语义搭配模式,通过概念-语义类型-语义类型组的两级映射,对语义述谓项进行类团划分。[结果 /结论]通过与常规共词分析方法比较,发现基于细粒度语义关系的共词分析能有效地揭示文本主题内容,利用UMLS语义网络资源能从语义学角度清晰准确地对语义共词网络进行类团划分。
[Purpose / Significance] Aiming at the common problems of co-word analysis, this paper proposes a method of constructing and analyzing co-word networks based on fine-grained semantic analysis. [Methods / Processes] With the help of SemRep, the concept of the source text and its semantic relation are normalized to extract and construct the semantic co-word network. Then, the content feature words are extracted by using the center of the nodes and the frequency of the edges as an index. UMLS semantics The network specifies the semantic collocation pattern, classifying the semantic description items by the two-level mapping of concept-semantic type-semantic type groups. [Results / Conclusion] By comparing with the common co-word analysis method, it is found that the co-word analysis based on the fine-grained semantic relation can effectively reveal the content of the text topic. Utilizing the UMLS semantic network resources can clearly and accurately analyze the semantic co-word network Class division.