WikiWars: A New Corpus for Research on Temporal Expressions

Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, pages 913–922, MIT, Massachusetts, USA, 9-11 October 2010. c⃝2010 Association for Computational Linguistics WikiWars: A New Corpus for Research on Temporal Expressions Paweł Mazur1,2 1Institute of Applied Informatics, Wrocław University of Technology Wyb. Wyspia´nskiego 27, 50-370 Wrocław, Poland pawel@mazur.wroclaw.pl Robert Dale2 2Centre for Language Technology, Macquarie University, NSW 2109, Sydney, Australia Pawel.Mazur@mq.edu.au Robert.Dale@mq.edu.au Abstract The reliable extraction of knowledge from text requires an appropriate treatment of the time at which reported events take place. Unfortu- nately, there are very few annotated data sets that support the development of techniques for event time-stamping and tracking the progres- sion of time through a narrative. In this paper, we present a new corpus of temporally-rich documents sourced from English Wikipedia, which we have annotated with TIMEX2 tags. The corpus contains around 120000 tokens, and 2600 TIMEX2 expressions, thus compar- ing favourably in size to other existing corpora used in these areas. We describe the prepa- ration of the corpus, and compare the profile of the data with other existing temporally an- notated corpora. We also report the results obtained when we use DANTE, our temporal expression tagger, to process this corpus, and point to where further work is required. The corpus is publicly available for research pur- poses. 1 Introduction The reliable processing of temporal information is an important step in many NLP applications, such as information extraction, question answering, and document summarisation. Consequently, the tasks of identifying and assigning values to temporal ex- pressions have recently received significant attention, resulting in the creation of mature corpus annotation guidelines (e.g. TIMEX21 and TimeML2), publicly 1See http://fofoca.mitre.org. 2See http://timeml.org. available annotated corpora (ACE,3 TimeBank4) and a number of automatic taggers (see, for example, (Mani and Wilson, 2000; Schilder, 2004; Hacioglu et al., 2005; Negri and Marseglia, 2005; Saquete, 2005; Han et al., 2006; Ahn et al., 2007)). However, existing corpora have their limitations. In particular, the documents in these corpora tend to be limited in length and, in consequence, discourse structure. This impacts on the number, range and variety of temporal expressions they contain. Ex- isting research carried out on the interpretation of temporal expressions, e.g. by (Baldwin, 2002; Ahn et al., 2005; Mazur and Dale, 2008), suggests that many temporal expressions in documents, especially news stories, can be interpreted fairly simply as be- ing relative to a reference date that is typically the document creation date. This phenomenon does not carry over to longer, more narrative-style documents that describe extended sequences of events, as found, for example, in biographies or descriptions of pro- tracted geo-political events. Consequently, existing corpora are not ideal as development data for systems intended to work on such historical narrations. In this paper we introduce a new annotated corpus of temporal expressions that is intended to address this shortfall. The corpus, which we call WikiWars, consists of 22 documents from English Wikipedia that describe the historical course of wars. Despite the small number of documents, their length …

Content truncated to 500 words