Microsoft has published a dataset of 100,000 question and answer pairs called Microsoft Machine Reading Comprehension (MS MARCO) that can help AI developers train their systems to interpret and answer questions. The data is comprised of anonymized real-world English-language searches on Microsoft’s Bing and Cortana virtual assistant. The goal of MS MARCO is to spur the development of AI systems that can answer questions as well as a human.
Training Machines to Answer Questions Better
Joshua New is a policy analyst at the Center for Data Innovation. He has a background in government affairs, policy, and communication. Prior to joining the Center for Data Innovation, Joshua graduated from American University with degrees in C.L.E.G. (Communication, Legal Institutions, Economics, and Government) and Public Communication. His research focuses on methods of promoting innovative and emerging technologies as a means of improving the economy and quality of life. Follow Joshua on Twitter @Josh_A_New.