Task Description


The evaluation has ended.


See the official paper of TempEval-3 task: http://www.aclweb.org/anthology/S/S13/S13-2001.pdf

or ArXiv paper http://arxiv.org/abs/1206.5333


The aim of TempEval is to advance research on temporal information processing, which could eventually help NLP applications like question answering, textual entailment, summarization, etc. TempEval-3 follows on from previous TempEval events, incorporating: a three-part task structure covering event, temporal expression and temporal relation extraction; the use of the complete set of TimeML temporal relations, that was simplified in previous editions; a 10-times larger dataset; and single overall performance scores, which allow the ranking of the participating systems in each task and also in general..


Task Organizers:

James Allen (james@cs.rochester.edu), University of Rochester
Leon Derczynski (leon@dcs.shef.ac.uk), University of Sheffield
Hector Llorens (hectorlm1983@gmail.com), University of Alicante
James Pustejovsky (jamesp@cs.brandeis.edu), Brandeis University
Naushad UzZaman (naushad@cs.rochester.edu), University of Rochester [Primary Contact]
Marc Verhagen (marc@cs.brandeis.edu), Brandeis University

Note: please subscribe to http://groups.google.com/group/tempeval


Contact Info


- James Allen, University of Rochester
- Leon Derczynski, University of Sheffield
- Hector Llorens, University of Alicante
- James Pustejovsky, Brandeis University
- Naushad UzZaman, University of Rochester/Nuance Communications [Primary Contact]
- Marc Verhagen, Brandeis University

email : naushad@cs.rochester.edu

Other Info


  • Program of SemEval 2013 (includes proceedings)
  • March 4, 2013. Evaluation data is released, find it in data and tools section.
  • February 26, 2013. Evaluation toolkit is updated. Please update from data and tools section.
  • September 3, 2012: Hand annotated Spanish training data is available in the data and tools section.
  • August 30, 2012: Silver data is available in the data and tools section. Human reviewed portion of the silver data will be available soon as additional gold data. Cleaned and improved AQUAINT and TimeBank corpus is already available as gold standard data.
  • March 26, 2012: Cleaned and improved AQUAINT and TimeBank corpus is available as gold standard training data for TempEval3. Download from data and tools section.
  • March 26, 2012: Toolkit released to evaluate the entity extraction (Task A and B) and relation performance (Task C). Download from data and tools section.