In today’s world, often called the digital era, large amounts of data are produced and analyzed every second, and thus pattern discovery from such data becomes inevitable. Data mining techniques work as powerful instruments for this purpose. In social sciences, nevertheless, few studies have been done regarding applications of data-driven and algorithmic methods. In this paper, the population census in 2011 was used and two indicators namely, educational attraction and occupational attraction were designed and implemented based on the model of Particle Swarm Optimization (PSO) algorithm in MATLAB software. As a result of applying these two indicators, the ranking score of 22 districts in Tehran was determined. According to the results, roughly 33% of the migration to Tehran is due to educational and occupational attractions (22% occupational, 11% educational), while other 67% of migrations is because of other factors. In terms of migration share, District 5 also displays the highest values in relation to educational and occupational attraction; however, District 21 indicates the lowest migration share and attraction. The findings show that applying intelligent algorithms within the data mining approach in the field of AI technology considerably improves the ability of such approaches in social science data analysis. In this respect, migration has been analyzed as a case study for illustration of this approach.