Businesses increasingly rely on patent valuation in order to evaluate their own position regarding intellectual property, as well as their competitor’s. Being able to statistically valuate patents is regarded by many professionals in the field as the ‘holy grail’ of patent landscaping. Statistical patent valuation is also of high interest to academia, who primarily use the results to better understand the patenting system.
In this project I studied and found new ways for improving statistical models for patent valuation. Several new value indicators based on patent citation data were proposed and tested against an estimate of patent value provided by inventors. One of the results shows that distinguishing between different types of citations helps to improve indicators. Other significant effects related to the technological fields that are cited or being cited from were also found.
The indicators proposed in this study can readily be applied in a business context to improve the predictive power of patent valuation models. Furthermore, the indicators shed new light on how value is derived from patents.
Download presentation: Improving statistical patent valuation indicators using citation data