面向查询意图的信息检索技术

用户兴趣和行为的多样性使得为不同用户提供更符合其查询意图的搜索结果成为一个具有挑战性的任务.Web 2.0下的社会标签是用户为他们感兴趣的网页等对象进行标注行为的结果，用户用标签来描述自己感兴趣的话题.这些标签不但代表着用户的兴趣，而且是对网页承载信息的最好揭示.提出了面向用户查询意图的标签推荐方法，旨在把能够体现用户真正查询意图的标签选择出来.标签作为对查询关键词的补充，不仅可以弥补用户短查询的缺陷，而且可以根据标签与网页上曾被标注过的标签间的关系，更准确地判断用户查询意图与网页内容之间的相关度，从而把更符合用户查询兴趣的结果排在靠前的位置上.实验结果表明，该方法比现有的其他方法更有效，这也说明社会标注对更准确地捕捉用户真实查询意图确实有重要作用.

Information Retrieval Techniques Based on Query Intention

Because of the diversity of users' interests and behavior, it is a great challenge to provide appropriate search results for different users. The social tags in Web 2.0 result from users' annotation behavior for pages that they are interested in. The user uses tags frequently to describe these topics, therefore the tags not only represent the user's query intention, but also reveal the Web information best. This paper presents a research for user-oriented query intention. It aims at adding the tags which can reflect the user's real query intention to the query. Tags as a supplement to the query keywords, not only can make up the defects of short queries, but also accord to the correlation between the query tags and Web content in sorting the related results. If the user just uses the keywords, the query log needs to be used to recommend high-quality tag as a supplement to the query keywords, and return more relevant results. Experimental results show that the proposed method is better in meeting the needs of users than other methods, making the search results more close to user requirements. This means that the social tags can make an important contribution to catching the query intention.