It is important to deliver appropriate services to requested users. In case of unavailability of a user requestedcomposite service, enforces the system to invoke service selection that involves choosing individual concrete services towards service composition. The services are selected based on two criteria: i) functional based and ii) nonfunctional based. The former entails selection of services based on functional property that the service is dedicated to do and the latter elite selection of services based on the QoS attributes such as reliability, availability, cost, and response time. Several population-based and swarm-based optimization algorithms are widely used for the process of web service selection. In this work, we employ a stochastic optimization algorithm called Self Organizing Migrating Algorithm (SOMA) and compare its performance with GA and PSO. The comparative study evidences that SOMA produces promising results and is therefore able to select user requested service in an efficient manner.