Classifier-Based Tense Model for SMT

Tense of one sentence can indicate the time when an event takes place. Therefore, it is very useful for natural language processing tasks such as Machine Translation (MT). However, the mapping of tense in MT is a very challenging problem as the usage of tenses varies from one language to another. Aiming at translating one language (source) which lacks overt tense markers into another language (target) whose tense information is easily recognized, we propose to use a classifier-based tense model to keep the main tense in target side consistent with the one in source side. Furthermore, we present a simple and effective way to help this model by expanding more phrase pairs with different tenses. Experimental results demonstrate our methods significantly improve translation accuracy. TITLE AND ABSTRACT IN ANOTHER LANGUAGE (CHINESE)