【摘要】 该研究旨在利用众包方法集成Web上的结构化信息，构建一个强大的知识库。传统模式匹配技术在处理网络表格的不完整性时存在局限性，特别是在发现不同列之间的语义对应关系方面。提出一种混合机器众包方法，旨在克服传统模式匹配技术的局限性。该方法基于理解表的语义，选择最有价值的列进行众包验证，并整合结果来推断其他列的概念，以实现表格匹配。该研究为Web数据集成提供了有效解决方案，为知识库的构建奠定了坚实基础。更多还原

【Abstract】 This study aims to integrate structured information on the Web using crowdsourcing methods to build a robust knowledge base. Traditional pattern matching techniques have limitations in dealing with the incompleteness of web tables, especially in discovering semantic correspondences between different columns. This article proposes a hybrid machine-crowdsourcing method to overcome the limitations of traditional pattern matching techniques. The method is based on understanding the semantics of tables, selecting the most valuable columns for crowdsourcing verification, and integrating the results to infer concepts of other columns to achieve table matching. This research provides an effective solution for web data integration and lays a solid foundation for knowledge base construction.更多还原