Abstract
In the recent years, social networks have played a strategic role in the lives of many companies. Therefore several decision makers have worked on these networks for making better decisions. Furthermore, the increased interaction between social networks and web users has lead many companies to use a data warehouse to collect information about their fans. This paper deals with a multidimensional schema construction from unstructured data extracted from social network. This construction is carried out from Facebook page in order to analyze customers' opinions. A real case study has been developed to illustrate the proposed method and confirming that he social network analysis can predict chance of the success of products.