Research And Implementation Of Parallel Query Technology Based On Semantic Caching

With the rapid development of information technology, the scales of data that the computer processed keep on growing. After many years, many application systems have stored massive data in their databases in order to analyze. Such massive information database system annual data scale always achieves TB level. Face such massive information, How to manage these data effectively and improve query performance becomes the focus of current research.Query processing present excellent performance in such massive information database system. In particular, the statistic query which system often process always need much time. Thus, how to optimize statistic query and shorten respond time has become an important aspect of excellent performance for the database system.After analyzing of various optimize query solutions sufficiently, basing on several characteristics of Massive Information Database System, such as, massive information, many statistic queries and various applications, the thesis proposes a mid-tier semantic caching architecture that suit statistic applications in massive information database. The realization of query processing about semantic caching is the thesis emphases.Through the studying of mid-tier semantic caching and characteristics of massive information statistic applications, the thesis presents the mid-tier semantic caching architecture and organization of semantic caching. Based on these, the thesis proposes a mid-tier semantic caching query processing algorithms and given its realization.Finally, the thesis constructed several performance test cases on a large transaction process system to analyze the performance of the mid-tier semantic caching. These tests show that the scheme that the thesis proposed enhances the performance of massive information statistic applications markedly, and can fulfill its requirement.