| 摘要: |
| 为了增强关系数据库中的关键字搜索查询结果,考虑了多表之间以及元组之间的语义关系,提出了一种语义评分函数.该语义评分函数不仅涵盖了当前的评分思想,并且加入新指标来衡量查询结果与查询关键字之间的相关性.基于该评分函数,提出两种以数据块为处理单位的Top-K搜索算法,分别为BA(blocking algorithm)算法和EBA(early-stopping blocking algorithm)算法.EBA在BA基础上引入了过滤域值,以便尽早终止算法的迭代次数.最后实验结果显示语义评分函数保证了 |
| 关键词: Top-K 关键字搜索 关系数据库 信息检索 语义相似度 |
| DOI： |
| 分类号: |
| 基金项目:Supported by the Program for New Century Excellent Talents in University of China under Grant No.NCET-06-0290 (新世纪优秀人才计划); the National Natural Science Foundation of China under Grant No.60503036 (国家自然科学基金); the National High-Tech Research and Development Plan of China under Grant Nos.2006AA09Z139, 2007AA01Z192 (国家高技术发展计划(863)); the Fok Ying Tong Education Foundation under Grant No.104027 (霍英东优选资助课题); the Key Laboratory of Data Engineering and Knowledge Engineering for the Ministry of Education, Renmin University of China (中国人民大学数据与知识工程教育部重点实验室开放课题) |
|
| A Top-K Keyword Search for Supporting Semantics in Relational Databases |
|
WANG Bin,YANG Xiao-Chun,WANG Guo-Ren
|
| Abstract: |
| In order to enhance the search results of keyword search in relational databases, semantic relationship among relations and tuples is employed and a semantic ranking function is proposed. In addition to considering current ranking principles, the proposed semantic ranking function provides new metrics to measure query relevance. Based on it, two Top-k search algorithms BA (blocking algorithm) and EBA (early-stopping blocking algorithm) are presented. EBA improves BA by providing a filtering threshold to terminate iterations as early as possible. Finally, experimental results show the semantic ranking function guarantees a search result with high precision and recall, and the proposed BA and EBA algorithms improve query performance of existing approaches. |
| Key words: Top-K keyword search relational databases information retrieval semantic similarity |
