Particle Swamp Optimization (PSO) is an effective method for solving a wide range of problems. However, the most existing PSO algorithms easily trap into local optima when solving complex multimodal function optimization problems. In this paper, we explain the importance of PSO algorithm’s general purposes to optimize strategy which has various parameters that decide its conduct and viability in advancing a given issue. This study gives a rundown of the best selections of parameters for different advancement situations which should enable the specialist to accomplish better outcomes with less exertion. In this paper, we define an important in the PSO algorithm, the parameters and how to apply this algorithm to different type of datasets online and offline. Repeat Consumption Matrices Dataset has been tested with PSO algorithm to define the number of clusters using selection of parameters. The results are then having been compared with the Genetic Algorithm (GA) where the algorithms have been tested based on the number of cluster and DB index for each of the data. Based on the results, it is shown that the PSO algorithm gives good and efficient results compared to GA algorithm. The output of the research is a PSO-based clustering algorithm that can be used in data mining by providing accurate and robust results in clustering, which can be realized in web search engines and automatic doc-ument organization.