论文部分内容阅读
【目的】构建人才知识结构的自动抽取方法。【方法】基于网络信息采集技术、网页分析以及文本分词、语义网相关技术,构建基于网络环境的人才知识结构的自动抽取系统。【结果】实验验证了该系统的有用性,系统识别课程的整体准确率在95%以上,对半结构化文件,召回率在95%以上;对非结构化文件,部分文件召回率低于90%。【局限】课程识别的召回率受到词典库内容的制约。【结论】本方法能为人才知识结构研究提供有用的工具,符合构建人才知识结构的基本要求。
【Objective】 To construct an automatic extraction method of human knowledge structure. 【Method】 Based on network information acquisition technology, webpage analysis and text segmentation, semantic web related technologies, an automatic extraction system of knowledge structure based on network environment was constructed. 【Result】 The experiment proves the usefulness of this system. The overall accuracy of system identification course is more than 95%, and for semi-structured documents, the recall rate is above 95%. For unstructured documents, the recall rate of some documents is less than 90 %. [Limitations] The recall rate of course identification is subject to the content of the dictionary library. 【Conclusion】 This method can provide a useful tool for the study of human knowledge structure and conform to the basic requirements of constructing human knowledge structure.