IRETE: An improved RETE multi-entity match algorithm

This paper analyzes the processing characteristics and execution mechanisms of the classical RETE algorithm and points out its limitations when performing multi-entity match and multi-rule match. By employing several technologies such as nodes sharing, types preprocessing, and index-based searching optimization, we propose an improved version of the RETE algorithm, IRETE, which is tested under multi-entity and multi-rule circumstances to be a much more efficient match algorithm, at the cost of a bit of loss in network construction efficiency.