基于结构特征和元模型的中文表格语义分析方法

摘要

In light of the actuality of less comprehensive analysis on semantics of tables in Chinese the current technique is,in this paper we present the structural features and meta-model based semantics analysis method.By using the well-known first-order predicate calculus functions including Value (value function)and Num (number function),as well as combining pseudo code,we analyse the semantics of several kinds of the most common tables in Chinese and obtain their surface semantics,context semantics,restrict semantics between master table and sub-tables,additional semantics of subsidiary nature of table on table data,and relationship semantics behind the attribute value. Results of example verification indicate that the structural feature and meta-model are the effective way for analysing the semantics of tables in Chinese,the acquired information amount and categories of semantics exceed far than the existing methods.%针对现有技术对中文表格语义分析不够全面的现实，提出基于结构特征和元模型的语义分析方法。使用具有公知性的一阶谓词函数 Value（值函数）、Num（数量函数），结合伪编码，对几类最常见的中文表格进行语义分析，获取了中文表格的表面语义、上下文语义、主子表之间的限制语义、表附属性对表格数据的附加语义、属性值背后隐藏的关系语义。实例验证的结果表明，结构特征和元模型是中文表格语义分析的有效方法，获取的语义信息量和种类远超目前的方法。