Patent monitoring is an essential activity for the development of new technology. Only experts can analyze competing patents and each patent assessed takes a lot of time. Despite the ever-increasing performance of search engines, it is difficult to target
only relevant patents. Thanks to JMP Pro's text explorer, it is possible to build a predictive model that predicts whether a patent is relevant. This template uses title, abstract, and claims text. Learning the model is based on the notations provided by the expert on a sample of patents already appraised. Next, the model is provided with an expanded patent database. It is then possible to predict the relevance of a patent very effectively. The expert can thus focus on the few patents of interest.
