Patents for NLP Software: An Empirical Review

Abstract

Natural language processing (NLP) patents are one of the fastest growing niche segments in the technology patent marketplace. NLP technologies power the latest in Artificial Intelligence (AI) applications. For example, NLP technology supports Apple’s Siri, Amazon’s Alexa, and Facebook’s friend recommendation system. Yet, while the literature on software patents is visibly scaling – the literature specifically focused on NLP patents is non-existent.

This Article draws on a growing body of computational linguistics, intellectual property, and technology law scholarship to provide novel NLP patent analysis and critique. Further, this Article contributes the first empirical NLP patent review, including novel software descriptions, market modeling, and legal analysis relating to patent claims. First, this Article discusses the two main technical approaches to developing NLP software. Second, this Article models an evolving NLP patent dataset, offering economic insights, legal claims analysis, and patent valuation strategies.

Keywords: Natural Language Processing, Patents, Linguistics, Economics