The beauty of brimstone butterfly: novelty of patents identified by near environment analysis based on text mining

The novelty of a patent may be seen as those patterns that distinguishes it from other patents and scientific literature. Its understanding may serve for many purposes, both in scientometric research and in the management of technological information. While many methods exist that deal with a patent’s meta-information like citation networks or co-classification analysis, the analysis of novelty in the full text of a patent is still at the beginning of research and in practice a time-consuming manual task. The question we pose is whether computer-based text mining methods are able to identify those elements of such a patent that make it novel from a technological and application/market perspective. For this purpose we introduce and operationalize the concept of near environment analysis and use a three-step text mining approach on one of the patents nominated as finalist in the 2012 European Inventor Award contest. We demonstrate that such an approach is able to single out, content-wise in a near environment, the novelty of the patent. The method can be used also for other patents and—with adaption of the near environment analysis—for scientific literature.