This paper proposes a new evolutionary algorithm for life log data mining. The proposed algorithm is based on the particle swarm optimization. The proposed algorithm focuses on three goals such as size reduction of data set, fast convergence, and higher classification accuracy. After executing feature selection method, we employ a method to reduce the size of data set. In order to reduce the processing time, we introduce a simple rule to determine the next movements of the particles. We have applied the proposed algorithm to the UCI data set. The experimental results ascertain that the proposed algorithm show better performance compared to the conventional classification algorithms such as PART, KNN, Classification Tree and Naïve Bayes.