Eric Nyberg

Language Technologies Institute
School of Computer Science
Carnegie Mellon University
5000 Forbes Avenue
Pittsburgh, PA 15213

Office: GHC 6715
Phone: (412) 268-7281
ehn AT cs DOT cmu DOT edu


Deeper QA: CMU, Watson, and the Open Advancement of Question Answering

OAQA provides a foundational architecture and methodology for accelerating collaborative research in automatic question answering. By making long term commitments to an architecture that supports and standardizes simultaneous research experiments, the OAQA is accelerating the advancement of QA technologies. Pioneered by Eric Nyberg and David Ferrucci (IBM), OAQA combines an object-oriented software architecture with comprehensive metrics, measurement and error analysis at the system and module levels. OAQA also defines challenge problems which require different levels of performance along dimensions such as real-time speed, answer accuracy and answer confidence. More details can be found in the OAQA technical report.

CMU has been collaborating with the IBM DeepQA Group since 2007 to develop an open-source framework for OAQA. The initial OAQA architecture and data model were used to build systems for the TREC challenge problem and the Jeopardy! challenge problem. In 2010, Carnegie Mellon students created open-source QA systems for English and Chinese using the OAQA architecture and available open-source components from CMU and elsewhere. Following OAQA principles, teams of current students work to improve these systems as part of their ongoing research and coursework. Carnegie Mellon students Nico Schlaefer and Hideki Shima also contributed algorithms and code to IBM's Watson system, as participants in IBM's summer internship program.



Other Projects

  • Machine Reading: Automated reading to capture knowledge from text
  • JAVELIN: Open-domain Question Answering
  • UCR: The UIMA Component Repository
  • RADAR: Reflective Agent with Distributed Adaptive Reasoning
  • CAMMIA: Dialog Management for Mobile Information Access
  • HALO: Building a Digital Aristotle
  • KANT: Knowledge-based Machine Translation
  • Eric Nyberg's technical publications on machine translation (Google Scholar)
  • IIM: Visual Programming for Information Management
  • U.S. Patents issued for past work on Controlled Language,
    Machine Translation, and Assistive Technology
Last Updated 13-Feb-2011