基于Web挖掘的搜索关键词建议研究

发布时间：2018-12-13 17:50

【摘要】：随着信息爆炸时代的到来,海量的信息在满足人们对各种信息需求的同时,也造成了信息泛滥、信息迷航等问题。传统的搜索引擎利用查询扩展或关键词建议一定程度上解决了这个问题。但目前的关键词建议并不是面向用户兴趣推荐关键词,它推荐的关键词是大众关注的热点或者是与查询词有包含关系的词组,这一方面不能满足用户个性化需求,另一方面也间接的影响到了检索的查全率和准确率。基于此,本文深入开展了基于web挖掘的搜索关键词建议研究。论文首先研究和分析了现有web挖掘技术和搜索引擎技术,对web挖掘与搜索引擎技术的结合做了可行性分析。在此基础上,本文提出了基于web挖掘的搜索关键词建议方法,并建立了系统模型,对实现系统的关键技术进行了深入研究。系统采用了基于同义词扩展的源数据获取方法,改进了TF×IDF权重公式,将web使用挖掘和web文本挖掘技术相结合分别应用于系统的源数据处理和特征词聚类阶段,论文完成了基于web挖掘的搜索关键词建议系统的设计与开发,通过对系统的测试分析,结果表明：该系统功能设计合理,在面向用户兴趣提供个性化关键词建议功能上有较好的改进作用。最后,论文指出了在源数据获取、文本聚类算法和中文分词技术方面还有待进一步深入研究。
[Abstract]:With the arrival of the era of information explosion, mass information not only meets the needs of people for all kinds of information, but also causes information flooding, information confusion and other problems. Traditional search engines use query extensions or keyword suggestions to solve this problem to a certain extent. However, the current keyword recommendation is not oriented to user interest, but is a hot topic of public concern or a phrase with a relationship with query words, which can not meet the personalized needs of users. On the other hand, it also indirectly affects the recall and accuracy of retrieval. Based on this, this paper deeply develops the research of search keyword suggestion based on web mining. Firstly, this paper studies and analyzes the existing web mining technology and search engine technology, and makes a feasibility analysis on the combination of web mining and search engine technology. On this basis, this paper puts forward the method of search keyword suggestion based on web mining, establishes the system model, and deeply studies the key technology to realize the system. The system adopts the method of source data acquisition based on synonym extension, improves the weight formula of TF 脳 IDF, and applies web usage mining and web text mining to the source data processing and feature word clustering stage of the system, respectively. This paper has completed the design and development of the search keyword suggestion system based on web mining. The test and analysis of the system show that the system function design is reasonable. In the user-oriented interest to provide personalized keyword advice function has a better improvement. Finally, the paper points out that further research is needed in source data acquisition, text clustering algorithm and Chinese word segmentation technology.
【学位授予单位】：中南大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.3

【参考文献】