The Text Analysis Conference (TAC) is organized by the U.S. National Institute of Standards and Technology (NIST). It was initiated in 2008 and developed out of NIST's Text REtrieval Conference (TREC) and Document Understanding Conference (DUC). TAC's mission is to support research within the Natural Language Processing community by providing the infrastructure necessary for large-scale evaluation of NLP methodologies. TAC's primary purpose is not competitive benchmarking; rather, the emphasis is on advancing the state of the art through evaluation results.
A TAC cycle consists of a set tracks, areas of focus in which particular NLP tasks are defined. NIST distributes test data for each track; participants run their NLP systems on the data and return their results to NIST; NIST then pools the individual results, judges them for correctness, and evaluates the results. The TAC cycle culminates with a workshop that is a forum for participants to discuss their work and plan future tasks and evaluations.
This volume constitutes the workshop notebook for the TAC 2011 workshop, held in Gaithersburg, Maryland, November 14-15, 2011. TAC 2011 brings together three tracks - Knowledge Base Population (KBP), Recognizing Textual Entailment (RTE), and Summarization. Seventy-five teams submitted results to one or more of the TAC 2011 tracks. The teams represented both universities and companies and were based in 20 different countries and five continents, with the greatest participation from Asia (29), Europe (25), and North America (19).
The TAC 2011 workshop includes plenary talks, a poster/demo session, and discussion and planning sessions for TAC 2012. Participants may sometimes cite specific vendors and commercial products, but the inclusion or omission of a particular company or product implies neither endorsement nor criticsm by NIST. Any opinions, findings, and conclusions or recommendations expressed in the individual papers are the authors' own words and do not necessarily reflect those of the sponsors.
TAC 2011 is sponsored by NIST and the U.S. Department of Defense, with additional support for the RTE track from the PASCAL-2 Network of Excellence. Their sponsorship is gratefully acknowledged, as is the invaluable contribution of the TAC 2011 track organizers and track participants.
Hoa Trang Dang, NIST