异构数据集成系统中的语义缓存模型

Model of semantic cache in heterogeneous data integration system

主要研究语义缓存,区别于传统的元组缓存、页缓存等方式.语义缓存具有由客户的查询生成,并且被客户的查询所访问的特点.客户端缓存查询的谓词和结果,当客户发出新的查询时,通过对缓存数据的语义检查,可以快速地从缓存数据中获得部分或者全部的查询结果,而不一定将查询全部提交到服务器进行处理.签客户端的缓存数据对于提高C/S模式的系统效率起着非常重要的作用,调查了目前的语义缓存研究现状,并且结合所研究的系统特点,提出了一种语义缓存的模型.

We introduce a caching scheme which is called semantic caching. Unlike the traditional page caching, tuple cache etc., the semantic cache is defined and accessed based on the queries submitted by the client. Both the predicate and the resulting tuples of queries are cached. When a new query is requested, the partial or whole result may be obtained through query processing on the semantic cache. Only the rest query is submitted to server. Data Caching at local clients plays an important role in improving the performance of client-server systems. In this article, with investigating current researches about semantic cache and taking account of the character of our research, we propose a kind of model of semantic cache.