Patent value predictions using interpretable machine learning.

By applying machine learning and NLP techniques to the information contained in patent documents we will develop a model for the valuation of patents. Such information will help firms in optimizing their patent portfolio, or, in price-setting of patents for licensing and patent transfers. There are three different work packages, the first one regarding the white spaces of the business environment. Secondly, regarding the timing, this relates to how technologies evolve over time, the technology lifecycle and the timing of new innovations. The final work package relates to the valuation of technologies along the technology trajectory and how one extracts value from technology portfolios.