Paper: Temporal Relation Identification with Endpoints

ACL ID N10-3008
Title Temporal Relation Identification with Endpoints
Venue Proceedings of the NAACL HLT 2010 Student Research Workshop
Year 2010

Temporal relation classification task has is- sues of fourteen target relations, skewed dis- tribution of the target relations, and relatively small amount of data. To overcome the is- sues, methods such as merging target relations and increasing data size with closure algo- rithm have been used. However, the method using merged relations has a problem on how to recover original relations. In this paper, a new reduced-relation method is proposed. The method decomposes a target relation into four pairs of endpoints with three target rela- tions. After classifying a relation of each end- point pair, four classified relations are com- bined into a relation of original fourteen target relations. In the combining step, two heuris- tics are examined.