Machine Learning Classification of VC Contract Terms

Jessica Jeffers, Assistant Professor and David G. Booth Faculty Fellow

We seek to create new data on VC contracting, especially around the emerging area of impact investing. We plan to create a public facing website where VCs and entrepreneurs can assess the relative strengths of their contracts, across multiple dimensions, by uploading legal documents. Building off pilot data and a process we have already developed, we will leverage natural language processing and maching learning algorithms to evaluate the contract terms supplied. Participants will receive a contract score benchmarked against comparable documents and an opportunity to provide feedback on the scoring application. Researchers will receive three benefits from the process: (1) access to otherwide hard to obtain data about private market investments, (2) improvement of the research process through automation and incorporating feedback, and (3) an audience for market research from a growing repository of VC investors and entrepreneurs.