学术搜索是一种行业化的搜索引擎,因其缺乏个性化、智能化的服务,使得用户的学术文献检索效率低下,海量的数字学术资源得不到充分利用。本研究跨语言智能学术搜索系统的设计与实现,旨在让用户可以在尽可能短的时间内找到所需学术资源。系统的几个关键技术包括:研究混合语种文本的分词技术;研究基于机器翻译的跨语言信息检索;研究搜索结果聚类算法在不同语言文本上的性能差异问题;研究基于聚类的个性化信息检索方法以及交互式查询扩展技术。实验测试结果表明:系统具有较好的扩展性,能为用户提供良好的学术检索服务。
Academic search is a domain-oriented search engine,while the lack of personalized,intelligent services result in inefficiency of literature retrieval and insufficiency of massively digital academic resources.The design and implementation of the Cross-Language Intelligent Scholar System were explored to enable users to retrieve academic resources in the minimum time.Several key techniques of the system include: a cross-language academic search engine based on machine translation,researched on the performance difference of search results clustering algorithms used at different language text,researched on clustering-based personalized information retrieval and interactive query expansion approach.The experiments showed that our system had a better scalability,which could provide preferred academic search services for users.