TERSEO system has been developed at the Natural Language Processing and Information System Group at the University of Alicante (partially supported by the Spanish government, project TIN-2006-15265-C06-01 and the framework of the European Project QALL-ME FP6-IST-033860), and its name stands for Temporal Expression Recognition System applied to Event Ordering.

In a first step, TERSEO was developed as a knowledge based system, in order to automatically recognize and normalize temporal expressions in Spanish texts. The architecture is shown in the following figure:


After this, TERSEO was extended to other languages with the automatic creation of the temporal models for new languages, starting from previously defined ones, so as to overcome the problems that are inherent in the rule-based paradigm. The English model was obtained automatically from the Spanish one, through the automatic translation of the Spanish Temporal expressions into English using a set of Machine Translation resources. The resulting system was called MT-TERSEO. Our final aim is to obtain an architecture similar to Wordnet, with a central unit with the Normalization rules (TER-ILI unit) and the different Recognition rules for the different languages (which is language dependent) and relate them to the TERILI unit.


The system generates TIMEX2 annotations as specified in the 2005 TIMEX2 guidelines for English and Spanish.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License