INTEGRATED EVALUATOR EXTRACTED FROM INFRINGEMENT LAWSUITS USING BACK PROPAGATION NEURAL NETWORK

This study aims at the basis of patent law and proposes an integrated evaluator constructed by a revolutionary evaluation model. Regardless of stock performance or revenue generated by the enterprise, the damage award of a patent infringement lawsuit is deemed to be a legal value of a patent in view of the patent law. Effective samples are extracted from 4,289 patent infringement lawsuits retrieved in U.S. district courts of Delaware, California and Texas. 17 quantitative indicators for describing dimensions of patents are summarized for the Back Propagation Neural Network incorporated with the Factor Analysis, so as to build the evaluation model which generating the integrated evaluator. The evaluation model is validated to be feasible by error analysis and is accommodated to applications of patent transaction deal, patent licensing, hypothecation of intangible assets, shareholding by patent-based technologies, and venture capital, etc.