基于语义缓存的移动查询导出

Extracting Query Results from Semantic Cache

在移动环境下 ,客户缓存为提高客户 -服务器数据库系统的整体性能 ,特别是保证客户端数据可用性提供了有效途径 .该文针对如何从基于语义描述的缓存中导出当前查询 (部分 )结果的问题 ,研究了查询从缓存导出的充分条件 ,并在定义查询与缓存之间的精确匹配、包含匹配和相交匹配几种情况的基础上 ,给出缓存与查询、包含与相交匹配的判断条件和相应的算法 .基于该文的研究 ,查询可以充分利用本地语义缓存的内容 ,从而降低网络开销 ,加快响应时间 ,并支持移动客户断接时的数据访问 .

This paper first defines query related cache item, extended attribute set and derivability, which are important concepts in extracting query results from cache, and then gives the sufficient condition to derive query. In the second part, different match types between cache and query, including exact match, containing match and overlap match, are defined. When query and cache item are exact match, query predicate is the same as cache predicate. When they are containing match, which means query result are completely contained in cache, query predicate must be implied by cache predicate. Authors present sufficient condition and corresponding algorithm to determine the implication relationship between query and cache predicate. When query and cache item are overlap match, only part of query result is in the cache. Authors also define query trim to trim the part that can be answered from cache and show different trimmed results in different overlap situation. Finally, this paper gives an algorithm to determine a trimmable overlap.Based on the work in this paper, query results can be extracted from semantic cache. Thus, query can take full advantage of local cache, which greatly lowers network bandwidth cost, improves query response time and supports data access during mobile disconnection.