Within the SemEval-2013 evaluation exercise, the TempEval-3 shared task aims to advance research on temporal information processing. It follows on from TempEval-1 and -2, with: a three-part structure covering temporal expression, event, and temporal relation extraction; a larger dataset; and new single measures to rank systems – in each task and in general.
17 PAPERS • 2 BENCHMARKS
French TimeBank, a corpus for French annotated in ISO-TimeML.
6 PAPERS • 1 BENCHMARK
KRAUTS (Korpus of newspapeR Articles with Underlinded Temporal expressionS) is a German temporally annotated news corpus accompanied with TimeML annotation guidelines for German. It was developed at Fondazione Bruno Kessler, Trento, Italy and at the Max Planck Institute for Informatics, Saarbrücken, Germany. Our goal is to boost temporal tagging research for German.
4 PAPERS • 1 BENCHMARK
TimeBankPT is a corpus of Portuguese text with annotations about time. The annotation scheme used is similar to TimeML. TimeBankPT is the result of adapting the English corpus used in the first TempEval challenge to the Portuguese language.
Spanish TimeBank 1.0 was developed by researchers at Barcelona Media and consists of Spanish texts in the AnCora corpus annotated with temporal and event information according to the TimeML specification language.
3 PAPERS • 1 BENCHMARK
A set of basque documents annotated with EusTimeML - a mark-up language for temporal information in Basque.
2 PAPERS • 1 BENCHMARK
Catalan TimeBank 1.0 was developed by researchers at Barcelona Media and consists of Catalan texts in the AnCora corpus annotated with temporal and event information according to the TimeML specification language.
HengamCopus is a Persian corpus with temporal tags (BIO standard tagging scheme). This dataset was generated by applying HengamTagger (https://github.com/kargaranamir/parstdex) to a large number of sentences. There are two types of Persian text datasets included in these collections: formal ones (Persian Wikipedia and Hamshahri Corpus), and informal ones (Twitter and HelloKish). In the creation of HengamCorpus, to maximize the diversity of patterns for training and evaluation, they uniformly draw samples from sets of sentences of unique “temporal pattern profile”, presence/absence vector of different temporal patterns within the sentence.
1 PAPER • 1 BENCHMARK