A new method for the discovery of the best threshold value for finding positive or negative association rules using Binary Particle Swarm Optimization

In association rule mining most of former researches have worked on analytic optimizing method , but finding and specifying the advocate initiation limit influences on association rule mining's quality , which still is important hence this research wants to present a new algorithm for optimizing the analytic efficiency improvement including automatic analyze proper amount for initiation. Through former method this task had been performing based on positive rules but regarding that finding the negative ones were though for administrator, this research's privilege is that the initiation level automatically is analyzed for the first time; also it has high efficiency in large data base. Particle Swarm Optimization is observed for any particle's efficiency and as data turned in binary the advocate amount will be found. Results showed Particle Swarm Optimization could present better initiation level, and enhance the former algorithm's result a lot. Consequence will be comparing with Weka and Apriori..