Users all around the world widely publish contents and leave comments about products/services they experienced in social networking sites. With the emerging computational intelligence approaches, the data can be processed and transformed to valuable knowledge. In this study, we propose a methodology based on computational intelligence techniques for market analysis. In the proposed approach, first customers' comments are collected automatically, then sentiment analysis is applied to each message using artificial neural networks. At the third phase, themes of messages are determined using text mining and clustering techniques. In order to represent the outcomes of the computational intelligence, a real world example from GSM operators in Turkey is given.