In the BI world, the importance of facts is undeniable. Although, opinion plays a crucial role too due to its tendency to influence the decision making process. Nowadays, multiple unstructured information sources are free and open to use via theWeb, such as social networks. In fact, the amount of these data is in a constant growth creating a free gold mine that can be used to collect public opinion. Companies are aware of the importance of their clients' opinions; they do not ignore in any case the importance of the public opinion that is available on social networks since the standing and growth of these platforms. In this paper we propose a robust decisional platform that collects, stores, and exploits the textual data generated by users through social networks.Finally, we present a validation of our proposals in a real-life, it's about the telecommunication area at Oreedoo Algeria.