Learning and Reformulation for Efficient Multidatabase Retrieval

高效多数据库检索的学习和重构

• 批准号: 9313993
• 负责人: Craig Knoblock
• 金额: $ 25.93万
• 依托单位: University of Southern California
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1993
• 资助国家: 美国
• 起止时间: 1993-09-01 至 1997-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9313993&HistoricalAwards=false
• 关键词: Learning; Reformulation; Efficient; Multidatabase; Retrieval

9313993 Knoblock An important and difficult problem in Knowledge Management is how to efficiently retrieve information from distributed, heterogeneous multidatabase systems. This project addresses this problem by bringing to bear a richer set of knowledge about databases to optimize multidatabases queries. The idea is to use semantic knowledge of the contents of databases to reformulate queries into equivalent yet less expensive ones. Using this knowledge, the potential cost reduction is significantly more than can be derived from convention query optimization alone. Since the semantic knowledge required can be learned from any database, this approach is applicable to other Knowledge-intensive systems. ***

小行星9313993 如何有效地从分布、异构的多数据库系统中检索信息是知识管理中的一个重要而又困难的问题。 该项目通过引入更丰富的数据库知识来优化多数据库查询来解决这个问题。 其思想是使用数据库内容的语义知识来将查询重新公式化为等价但成本较低的查询。 使用这些知识，潜在的成本降低比仅从常规查询优化获得的成本降低要多得多。 由于所需的语义知识可以从任何数据库中学习，这种方法适用于其他知识密集型系统。 ***