I have a problem and I'm not sure how to approach it. I have a bunch of resolved cases, with the original problem description, and an identifier that points to the method used to resolve the case. I'd like to use Term Extraction on these resolved cases to relate the terms used in the original problem descriptions to the resolving method. So I'm thinking that the data modeling structure would consist of a case with the resolving method and a nested table of all the related terms. However, what I want to do is then take a new problem description, extract all the terms for the description of the problem description, and then use these terms to predict what the resolving method would be. What's the best approach to do so? Could I use Microsoft Association Rules to do so? Or could I use Clustering? I'm pretty much a newbie when it comes to Data Mining.