Text mining applied to patent mapping: a practical business case

Professional patent searchers are traditionally rather suspicious of the alleged “black box” effect inherently attached to intelligent software engines relying upon linguistic technologies for patent analysis and mapping. In this article, the authors propose that such prejudices can be overcome by setting a realistic business objective while experimenting with these new linguistic tools, as well as by applying serious methodology for validating the results of the analysis. The strengths and weaknesses of a particular text mining tool are assessed with reference to a practical business case in the field of packaging technology, and a comparison of the outcome of such an analysis with a traditional one, carried out using conventional patent classifications, is also described.