基于Agent的智能元搜索引擎调度与结果合成

发布时间：2018-05-15 10:23

本文选题：元搜索引擎 + Agent　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：大数据环境下,网络数据过载与用户需求提升,在信息覆盖、智能服务方面传统搜索引擎往往表现出许多明显劣势。针对这些方面,智能化的元搜索引擎被提出来解决以上存在的问题,并得到了广泛的关注。本文将Agent技术引入元搜索引擎,重点研究其中的成员搜索引擎的自适应调度策略与检索结果的动态合成方法。本文给出了基于Agent的智能化元搜索引擎调度及结果合成的设计与实现。首先明确了当前搜索引擎及元搜索引擎存在的不足之处并对相关的国内外研究现状进行了剖析,然后在介绍实现相关功能所需要的理论与技术,如Agent、成员引擎调度与检索结果合成等的基础上,提出了智能化的元搜索引擎体系结构。接下来重点研究了成员搜索引擎的自适应调度策略,构造了搜索引擎能力模型,提出了基于概念格的初始评估方法与用户点击行为反馈评估相结合的搜索引擎能力评估方法,并给出了具体的成员搜索引擎动态自适应调度过程的实现。随后重点介绍了元搜索引擎结果动态合成方法,设计了基于结果位置、搜索引擎权重及用户兴趣三者结合的一种改进的结果合成排序方法。最后对成员引擎的调度及检索结果合成进行了实验,测试结果表明,在基于Agent的智能元搜索引擎框架下,本文提出的调度策略能动态自适应的进行调度,选择最符合用户意图的成员搜索引擎,检索结果的动态合成排序方法能明显提高信息覆盖率与搜索相关程度,并且使用户的搜索效率得到了保证。
[Abstract]:Under the big data environment, the network data overload and the user demand enhancement, in the information coverage, the intelligence service aspect traditional search engine often displays many obvious disadvantages. In view of these aspects, intelligent meta-search engine has been proposed to solve the above problems, and has been widely concerned. In this paper, the Agent technology is introduced into the meta search engine, and the adaptive scheduling strategy of the member search engine and the dynamic composition method of the retrieval results are studied. This paper presents the design and implementation of intelligent meta-search engine scheduling and result composition based on Agent. Firstly, the deficiency of the current search engine and meta search engine is clarified, and the current research situation at home and abroad is analyzed, and then the theory and technology needed to realize the relevant functions are introduced. On the basis of agent, member engine scheduling and retrieval result composition, an intelligent meta-search engine architecture is proposed. Then, the adaptive scheduling strategy of member search engine is studied, and the search engine capability model is constructed. A new method of search engine capability evaluation based on concept lattice and user click-behavior feedback evaluation is proposed. The realization of dynamic adaptive scheduling process of member search engine is also given. Then, the dynamic composition method of meta-search engine results is introduced, and an improved ranking method based on the combination of result location, search engine weight and user interest is designed. Finally, the scheduling of the member engine and the composition of the retrieval results are experimented. The test results show that the scheduling strategy proposed in this paper can be scheduled dynamically and adaptively under the framework of intelligent meta search engine based on Agent. By selecting the member search engine that is most suitable to the user's intention, the dynamic composition sorting method of retrieval results can obviously improve the degree of information coverage and search correlation, and ensure the search efficiency of users.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.3

【相似文献】