Research On Pattern Match Algorithm In Rule Engine And Application For Rule Engine

In recently years, software application industry is dedicated to apply rule-based expert system technology to enterprise service, which spurred the wild interests in rule engine systems. By separating business rules from the specific enterprise application, the rule engine system can significantly reduce the underlying complexity of various applications as well as helping to achieve good system flexibility.In this work, we focus on the research on the pattern match algorithm in rule engine, and resolve to improve Rete algorithm from both functionality and integration. When or logic, not logic and existential quantification are introduced to enhance rule's representing ability, Rete algorithm will become complicated. In order to reduce the complexity, this work proposes a rule decomposition algorithm and initial fact strategy to improve Rete algorithm from the perspective of functionality. By using common objects instead of facts managed by rule engine API, shadow proxy strategy is introduced to provide easy integration for Rete algorithm. This paper proposes a novel architecture of the rule engine system, and implements a prototype of the pattern match system in rule engine.This paper provides C++ rule engine general interface for the reference of system development engineers and rule engine vendors. This paper also proposes a solution on the integration of rule engine into application system as the following three steps:first constructing the data model and condition model to describe the dataset that will be inserted into the rule engine by application system, then revising the condition part in the original application system by iterative revision strategy, finally invoking rule engine according to the order presented in this work. In the end of this paper, we verify the integration solution and test the pattern match system by embedding rule engine into context billing system.