Electricité de France, the French electricity provider is moving into new markets, and needs to know very well its clients and tracks their satisfaction. Thus, different kinds of text mining tools were tested in order to analyze a huge quantity of heterogeneous textual documents, including mails, open-ended customer survey questions, discussion forums, and comments contained in large databases concerning customer contacts. In this context, it seemed unavoidable to build a test grid, in order to facilitate the comparison between different tools. This paper describes the test grid carrying out, the software selection, the way the evaluation of four tools (Alceste, SAS Text Miner, TEMIS Insight Discoverer and SPAD/CRM) was achieved and the results.
