Temporal Expression Recognition and Normalization Task

In the Temporal Expression Recognition and Normalization task, systems are required to recognize the Temporal Expressions occurring in a text by identifying their extension, and to normalize them, i.e. give a representation of their meaning by assigning values to a pre-defined set of attributes.

The task will be based on the TIMEX2 standard, with some adaptations to Italian. The Italian Content Annotation Bank (I-CAB) Version 4.1, developed by FBK and CELCT, will be used both for development data (335 news stories, for a total of 113,000 words) and test data (190 news stories, for a total of 69,000 words). I-CAB is distributed upon acceptance of the agreement terms for a free research licence. This task consists of two subtasks:

  • Temporal Expression Recognition
  • Temporal Expression Recognition and Normalization

Participants may choose to partecipate in any of the two subtasks.

Organizers Valentina Bartalesi Lenzi (CELCT, Trento, Italy - bartalesi[at]celct.it) and Rachele Sprugnoli (CELCT, Trento, Italy - sprugnoli[at]celct.it)

Data Distribution The I-CAB Corpus Version 4.1 (used for EVALITA 2007) is freely available for research purposes upon acceptance of a license agreement: obtain I-CAB 4.1 Contact: Manuela Speranza (manspera[at]fbk.eu)

Detailed guidelines

Trial examples: TE recognition and TE recognition and normalization

ACE-2007 scorer (modified for EVALITA 2007)

APF DTD

Annotation Report