基于概念格的查询扩展系统及建格算法研究-模式识别与智能系统专业论文.docxVIP

THE THE RESE根CH OF QUERY EXf!ANSION SYSTeM BASED ON CONCEPT I，ArrICE AND CONSTRUCⅡON ALGORITHM ABSTRACT With the running up of Wreb information，its quite di伍cult for users to look for information i11 vast information ocean，and information retrieval is just the new technology which appears to solve this problem． At the same time．user’S request to information retrieval iS higher and higher,SO quickness，exactness have become the weighing criterion of information retrieval system’S performance．However,its quite difficult for all the users to describe their retrievalintent exactly．The query which users input into the system iS always inconsistent with their actualintent． So，how to expand query accurately has been thought much of by more and more researchers． ne paper brings forward a query expansion method based on concept lattice．It processes part of the result documents generated by the user’S query through concept lattice technol095 and locates the query in the generated concept lattice．After that，it expands the query automatically and interacts with the user by the concept lattice diagram．The user can derive of the navigation and expand query manually,which can avoid the problem that result is worse and worse in some situmion．Besides the above．the system can recommend some retrieval direction to the useL The experiment results indicate that the query expansion system based on concept 1attice is advantaged for the users to describe their queries accurately,and it improves the information retrieval accuracy,and it also shows US the individually info·rmation retrieval． KEY WORDS：information retrieval，query expansion，concept lattice， lattice construction 创新性声明本人声明所呈交的论文是本人在导师指导下进行的研究工作及取得的研究 创新性声明 本人声明所呈交的论文是本人在导师指导下进行的研究工作及取得的研究 成果。尽我所知，除了文中特别加以标注和致谢中所罗列的内容以外，论文中不 包含其他人已经发表或撰写过的研究成果，也不包含为获得北京邮电大学或其他 教育机构的学位或证书而使用过的材料。与我一同工作的同志对本研究所做的任 何贡献均已在论文中作了明确的说明并表示了谢意。 申请学位论文与资料若有不实之处，本人承担一切相关责任。 本人签名： 日期： 邱!呈，!! 关于论文使用授权的说明 学位论文作者完全了解北京邮电大学有关保留和使用学位论文的规定，即： 研究生在校攻读学位期间论文工作的知识产权单