Chinese Temporal Relation Resolution Based on Chinese-English Parallel Corpus

This paper investigates the use of parallel corpus for the annotation of temporal objects -- EVENT-type and TIMEX3-type elements in Chinese text. We propose a TRR engine based on parallel corpus in Chinese text employing existing tools for the temporal recognition and normalization of English to annotate the English portion of English-Chinese bitexts, and automatically project these annotations to the Chinese text, guided by word alignment. Through the method, Chinese Time Bank has been constructed, which lays the foundation for the further TRR research based on Time ML in Chinese text.