Research And Implementation Of The Rule-based View Of Data Integration Environment Monitoring Mechanism

This paper analyses and classifies the existent data deriving technologies in data integration systems in order to show some defects existing in these technologies. For instance, they are unfit for multi-sources data integration problem, can not support view object data deriving and provide few personalized data deriving requirements, etc. To solve these problem, this paper introduces an EGA rule system based Information Sources Monitor Intelligent Agent(ISMIA) method. This method is able to satisfy the needs that support multi-sources data deriving and send view incremental data with as few as possible redundant data to target database initiatively according to data deriving demands.Information Sources Monitor Intelligent Agent consists of transaction log analyzer, EGA rule system and view incremental data computing and sending components. The ISMIA's instance running on each information source can accept monitor commands coming from target database and then initiatively sends view incremental data to target database according to surrounding situation's changes without outside direct intervention.Afterwards, this paper focuses on the design of the special EGA rule system used to control view incremental data computing and sending in the ISMIA. We define the EGA rule language and give the event triggering, condition evaluation and action execution methods. To evaluate rule condition we introduce a new discrimination network structure called C-Network that is improvement of the widely known Rete algorithms. Algorithms for performing pattern matching using the C-Network to see if manipulations to view's base table would cause data changes in the view are given. It provides a new method to analyze "irrelevant update" issues.Finally, we discuss the implementation of special EGA rule system in the ISMIA..