基于语义的科技文献检索技术分析-analysis of scientific and technological literature retrieval technology based on semantics.docxVIP

摘要科技文献检索是科技研究不可或缺的一项工作，随着电子文献资源的剧增， 通过网络来检索文献，是目前最为方便的检索途径。目前Internet上用户使用的信 息检索模式有基于关键词、基于全文检索技术等方式，但是这些模式在面对快速 增长的信息时，往往返回的查询结果列表过于庞大，或者返回的结果没有满足用 户对信息内容的需要。为此，应当出现更有效的、准确的信息检索技术解决当前 的检索问题。基于关键词的检索技术和检索方式严重地割裂了字、词间的语义关联。加之 自然语言中一义多词、一词多义现象的广泛存在，导致用户查询获得的检索结果 要么包括太多的无关信息，要么返回结果太少。因此，难以满足用户对于文献信 息资源查询的需求。针对这种情况，本文提出了一种基于语义的科技文献检索系 统方案。该方案通过建立领域本体库,为文献检索在特定领域提供了语义上的支 持，再通过本体与文献数据库的映射得到语义索引文件，最后在此基础上建立了 基于本体的科技文献检索系统，通过实验证明该原型系统能有效提高论文文献的 检索性能，弥补了传统检索方法的不足。关键字：统一检索，科技文献检索，语义，本体The Studies of Scientific and Technical Literature Retrieval Technique Based on SemanticAbstractScientific and technical literature search is essential for scientific and technological research work. With the increase in electronic literature resources, using the network to search the literature is the most convenient way to search. At present, the mode of information retrieval used by Internet users has many ways: based on keywords, based on the full-text search technology, and so on. However, when these methods facing the rapid growth of information, the search-result lists are often too large, or the results of the return do not meet users’ information needs. Therefore, a more effective and accurate information retrieval technology should appear to solve the problem of the current search.The search technology and method which based on key words have seriously fragmented the word and semantic association between the words. In addition, a multi-word meaning and polysemy are widespread in the natural language, which result in search results. These results which users query are very unsatisfactory, including too much irrelevant information or few results. So it is difficult to satisfy the users’ demand for search for information resources and literature. To address this situation, this paper proposes a scientific and technical literature retrieval system which based on semantic method. First, build a domain ontology database for supporting literature search in specific areas of the semantic in