Abstract: At present, time expression recognition plays an important role in information extraction, ctuestion answering,summary generation, topic detection and tracking. Then, it has become one of the most important tasks in natural language processing. This article mainly studies two tasks of TempEval-2:the temporal expression recognition in one sentence and the temporal type identification. This paper proposes a method based on rule base,which is generated by "Unit of time" based POS tagging to indentify the time expression, and a method based on ME model to label the type of time expression. The experiments show that the accuracies of the two tasks are respectively 85.16% and 93. 02%.

Key words: Temporal expression, Rule, Time expression type, Maximum entropy model,Natural language processing