Predicting citations to biotechnology patents based on the information from the patent documents

Abstract

The purpose of this study is to develop a simple and robust model for predicting citations to a patent based on the information from the front page of the patent documents. The number of citations received is frequently used as an indicator of the value and importance of a patent. However, it takes a long time for a patent to accumulate a large number of citations from later patents. Highly cited patents may well be very old and accordingly may not represent cutting-edge technology. If we can predict the pattern of citations to a patent right after it is granted, the tradability of patents will be greatly enhanced. This paper provides a simple regression model to predict citations to biotechnology patents from the front pages of patent documents. The model can be used as a supplementary evaluation tool in mergers and acquisitions, strategic technology planning, valuation of high-tech firms and R&D performance evaluation.