【作者】 李雪；

【导师】 杜军平；

【作者基本信息】 北京邮电大学 ， 计算机技术（专业学位）， 2013， 硕士

【摘要】 本论文对旅游个性化搜索进行了研究,建立了旅游领域用户兴趣模型和主题实例(SID)模型,改进了查询扩展算法,建立了北京旅游信息知识库,分析了其距离属性、主题属性,扩展了概念属性和实例属性,实现了原型系统,满足了不同用户的需求。论文完成的主要工作如下：(1)针对传统的用户兴趣模型的单一性,本文提出了结合用户行为信息和景点区域权重的的个性化用户兴趣模型,给出了该模型的构建框架,设计了面向不同用户的景点排序算法,并将其应用于个性化搜索中,实现了精准搜索的查询推荐功能和模糊搜索的查询扩展功能,并通过实验验证了该模型的有效性。(2)改进了传统的关键词查询的单一模式,根据旅游领域特征,设计了相关度的互动模式,可准确定位用户搜索需求。实现了SID模型及算法,通过计算主题与主题相关度、主题与实例相关度以及实例与实例相关度,丰富了旅游领域知识信息库。改进了传统的搜索模式,并通过实验验证了该模式的有效性。(3)基于建立的用户兴趣模型和SID模型,改进了旅游领域搜索的查询扩展功能。通过建立双模型的扩展词集,提高了用户搜索覆盖率。根据景点优先权和景点相关度大小,对搜索结果进行了再排序展示,提高了查准率。(4)设计并实现了基于双模型的旅游个性化搜索系统,建立了知识库管理模块,该模块将景点优先权和主题相关度相结合,为搜索模块提供服务。完成了个性化搜索功能,结合用户输入的关键词和相关度,完成了个性化搜索的精准搜索、模糊搜索及推荐功能。本论文实现了一个旅游个性化搜索原型系统,提供全面的旅游信息搜索服务,帮助用户在浩如烟海的旅游信息中找到和需求相匹配的信息。更多还原

【Abstract】 This thesis mainly studies user Interest model in relevant fields and accomplishes subject relation model. Based on it query expansion technique is improved and tourism personalized search system is realized. The system takes the travel information of Beijing as knowledge base, analyses its distance properties, subject attribute and extends the concept attribute and instance attribute, which fully satisfies the different requirements of the users.(1) In the traditional system, user interest model is simple, this thesis presents a personalized model that combined with user behavior information and area-weight on attractions, and gives the model frame, designs sort algorithm, and applied to personalized search, so as to realize the query recommend while precise search and query expansion function with fuzzy search, in the end, it proves the validity of the model and the accuracy through the experiment.(2) To improve the single mode in traditional system, in which user can only input keywords, this thesis provides an interactive mode. According to the characteristics of tourism area, this thesis realizes the SDL model and relation algorithms, calculates the relation of the subjects, the concepts and the instances. We can enrich knowledge information database in the tourism domain, and facilitate the query expansion and search in later.(3) Based on the user interest model and the SDL model, this thesis makes the improvements on query expansion skills in the field of tourism search by extending the keywords set with the double model, displays the user’s demand information comprehensively. This way improves the coverage. According to the attentions priority and attentions relation, this system sorts the searching results again, which improve the precision ratio.(4) This thesis designs and realizes the travel personalized search system based on the double model,and designs the knowledge database management module, the module combines the attractions priority and subject relation, provides services for the upper searching module. The system completes the personalized search function, which includes precise search, fuzzy search and corresponding recommend results.The thesis implements a complete personalized search system, provides accurate comprehensive search function, makes full use of the characteristics of the domain knowledge information, combined with the user information and user behavior, launches a semantic analysis, and expands the searching keywords set.更多还原