TempCourt: evaluation of temporal taggers on a new corpus of court decisions

• DOI: 10.1017/s0269888919000195
• Publication Date: 2019-12-17T09:08:21Z
• Authors: María Navas-Loro; Erwin Filtz; Víctor Rodríguez-...; Axel Polleres; Sabrina Kirrane
• Keywords: Domain Adaptation

Abstract The extraction and processing of temporal expressions (TEs) in textual documents have been extensively studied several domains; however, for the legal domain it remains an open challenge. This is possibly due to scarcity corpora particularities found that are highlighted this paper. Considering pivotal role played by information when comes analyzing cases, paper presents TempCourt, a corpus 30 from European Court Human Rights, Justice, United States Supreme with manually annotated TEs. contains two different annotation sets adhere TimeML standard, first one capturing all TEs second dedicated relevant case under judgment (thus excluding dates previous court decisions). proposed gold standards subsequently used compare ten state-of-the-art cross-domain taggers, identify not only limitations taggers but also standard applied documents. Finally, identifies need resources adaptation existing tools, specific guidelines can be adapted types