Long-Distance Time-Event Relation Extraction

Abstract

This paper proposes state-of-the-art models for time-event relation extraction (TERE). The models are specifically designed to work effectively with relations that span multiple sentences and paragraphs, i.e., inter-sentence TERE. Our main idea is: (i) to build a computational representation of the context of the two target relation arguments, and (ii) to encode it as structural features in Support Vector Machines using tree kernels. Results on two data sets – Machine Reading and TimeBank – with 3-fold cross-validation show that the combination of traditional feature vectors and the new structural features improves on the state of the art for inter-sentence TERE by about 20%, achieving a 30.2 F1 score on inter-sentence TERE alone, and 47.2 F1 for all TERE (inter and intra sentence combined).