基于图形化定制的语义搜索系统的设计与实现

发布时间：2018-03-18 03:13

本文选题：语义搜索　切入点：本体　出处：《山东大学》2012年硕士论文　论文类型：学位论文

【摘要】：随着Web以及相关技术的高速发展,信息爆炸已经出现在当今的互联网上。搜索引擎是Internet上获取信息的最有效的工具。目前互联网上的文档大都是无结构或者是半结构的,计算机无法理解文档的语义信息。因此搜索引擎的查准率和查全率尚难以满足用户的实际需求,与此同时,已有的搜索引擎在进行领域内的某些查询时也显得不够智能化。本体和语义技术为改善Web搜索效果提供了新的思路,对使用基于语义的搜索技术提高信息检索的精确率的研究已取得了一定的成果,针对语义检索的方法开始出现,比如基本概念定位和复杂约束查询。语义搜索是将搜索引擎同语义网的相关技术结合来改善查询的结果。通过对语义Web的相关概念、相关开发技术(特别是Jena、Pellet、Protege-OWLAPI三个重要开发框架)以及对搜索引擎系统开发流程的研究,并借助Protege本体编辑工具和Graphviz绘图工具,本文设计并实现了一种结合SVG和AJAX技术图形化定制用户查询的语义搜索系统,使使用者能够直观的定制满足自己需求的查询。该系统使用web2.0中的核心技术Ajax去改善查询定制的交互性和用户友好性。系统可实现客户端上基于SVG的图形完成概念的定位以及像查询类别添加约束的过程。服务器接受用户的请求然后推理出基于布尔逻辑的概念实例,返回符合要求的相关网页文档。本系统通过设计如下八个模块：本体持久化模块,前台图形化定制页面模块,服务器控制器Servlets模块,SVG绘图功能模块,限制输入Suggest模块,树型视图处理模块,OWL本体解析模块,RDQL查询模块,实现了六项功能：本体类别层次浏览,设置查询类别,为类别添加关联对象属性,为类别添加关联数据属性,类别实例化,查询实例集合。实验结果表明,基于图形化定制的语义搜索系统相比传统的基于关键字检索的搜索引擎,能有效的改善查准率和查全率。
[Abstract]:With the rapid development of Web and related technologies, information explosion has appeared on the Internet today. Search engine is the most effective tool for obtaining information on Internet. At present, most documents on the Internet are unstructured or semi-structured. The computer cannot understand the semantic information of the document. Therefore, the precision and recall rate of the search engine is still difficult to meet the actual needs of the user, at the same time, The existing search engines are not intelligent enough for some queries in the field. Ontology and semantic technology provide a new way to improve the efficiency of Web search. Some achievements have been made in improving the accuracy rate of information retrieval by using semantic search technology. Semantic search is a combination of search engine and semantic Web technology to improve the result of query. Through the research on the related concepts of semantic Web, the related development technology (especially the three important development frameworks of Jena pellet pellet protege-owl API) and the development process of search engine system, and with the help of Protege ontology editing tools and Graphviz drawing tools, This paper designs and implements a semantic search system which combines SVG and AJAX technology to customize user query graphically. The system uses Ajax, the core technology in web2.0, to improve the interactivity and user-friendliness of query customization. The system can realize the completion of graphics based on SVG on the client side. The server accepts the user's request and deduces a conceptual example based on Boolean logic. This system designed the following eight modules: ontology persistence module, foreground graphical custom page module, server controller Servlets module, SVG drawing function module, limited input Suggest module, the system is designed as follows: ontology persistence module, foreground graphic customization page module, server controller Servlets module and Suggest module. Tree view processing module / owl ontology parsing module / RDQL query module, which implements six functions: browse ontology class hierarchy, set query category, add associated object attribute for category, add associated data attribute for category, instantiate category. Query the collection of instances. The experimental results show that the semantic search system based on graphical customization can improve the precision and recall effectively compared with the traditional search engine based on keyword retrieval.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP311.52

【参考文献】