A few years ago, three research groups participated in an audacious experiment called Project Halo: (manually) converting the information contained in one chapter of a high school chemistry textbook into knowledge representation statements, and then having the knowledge representation system take the high school AP exam.systems passed, albeit at a relatively low level of performance.ally?e October, several projects have taken up this challenge, or aspects of it.ng project at ISI, drawing part-time participation of experts in NLP and KR&R, addresses the problem from the perspective of NLP. Balysis and preparation, we parse the Chemistry textbook and then convert the results into very shallow sserted to a knowledge base.ill in progress, has two aspects. we apply questions to the system at various levels (text-only, knowledge level without inference, the latter with inference), and compare performance.groups) compare the systemb e groups.rojects are merely pilot studies, they nonetheless are likely to generate some interesting conclusions regarding the gap between what automated systems can deliver and what human knowledge engineers deem necessary, in the fascinating endeavor of learning by reading.
Collaborators: Jerry Hobbs, Chin-Yew Lin, Patrick Pantel, Hans Chalupsky, and students; USC Information Sciences Institute, University of Southern California