The IMAGIWEB ANR project is focused on studying the images portrayed on the Internet based on the detection of opinions. Two case studies were identified: (1) the first is designed to meet the analytical needs of researchers in policy science through data from Twitter during the 2012 presidential campaign ; (2) the second is to allow the French company EDF to be able to assess the public's opinion on safety, employment and prices from blog posts. Here we present a feedback on the use of OLAP (OnLine Analytical Processing) for textual data on these two case studies, demonstrating the interest for the project members.