Tools and techniques for recognizing temporal expressions from text: A survey

Temporal expressions are words that convey temporal information such as a point, duration or recurrence in time. These expressions can be used in extracting the temporal information or in time-related question answering, i.e. for answering “When” type of questions. Recognizing these temporal expressions from a chunk of text, by automatically annotating them is an increasing area of research. The temporal expressions have to be normalized. In other words, they have to be converted to some standard form in order to become understandable by algorithms. News articles have been taken up as the corpus for almost all of the recognition systems developed so far because these contain a plethora of temporal expressions. Using the manually annotated corpus, also called gold standard for computing the precision and recall values of available methodology, has been the technique to measure the accuracy. In this paper we present the various tools and methodologies adopted or developed in order to recognize the temporal expressions.