装备采办知识搜索引擎关键技术研究

摘要

研究并设计了装备采办知识搜索引擎系统结构，包括搜索模块、索引模块、检索模块、查询扩展模块和结果聚类模块。就装备采办知识管理搜索关键技术查询扩展模块和结果聚类模块，提出了相似性度量的渐进直推式支持向量机（sim-ilarity measurement progressive tranductive support vector machine ，SM PTSVM ）查询扩展模型，以及基于人工免疫算法的搜索引擎结果聚类算法。实验结果表明，该模型和算法提高了搜索结果的查全率、查准率及搜索结果的平均聚类精度。%This paper researches and designs the equipment procurement knowledge searching en-gine system structure .The system structure has five modules ,including scouting module ,indexing module ,searching module ,inquiring extension module and clustering module .About extend module and clustering module ,it proposes similarity measurement progressive tranductive support vector ma-chine (SMPTSVM) inquiring extension model and search engine results clustering algorithm based on artificial immune algorithm .The experimental results show that this method can improve the recall ratio ,precision and the average clustering precision of the search results .