Abstract:Innovations across a multitude of technologies and pathways will be necessary to substantially reduce greenhouse gas emissions from the transportation sector. This paper explores the use of Natural Language Processing (NLP) alongside machine learning algorithms to assess shifts in biofuel technology innovation. The full background text of over 755,000 patents from the U.S. Patent and Trademark Office patent database has been analyzed and classified using the Stanford NLP Classifier. This process provided a clean set of biofuel patent data compared to what has previously been done in the literature, and research and development (R&D) trends from this analysis were explored in the context of biofuel policies. The authors found that the primary U.S. policy affecting biofuels, the Renewable Fuel Standard, seems to have had limited impact on promoting innovation and patenting activity for 2nd generation fuels, but may have played a large role in encouraging the proliferation of entrepreneurial firms into the market. The authors conclude that, at this time, policies should be directed toward establishing networks, supply chains, and technology and knowledge sharing across firms already operating in this space to share learning and promote targeted technology outcomes.

Note:Presented at Transportation Research Board 94^th Annual Meeting, Washington, DC, January 11-15, 2015