CRF based Approach for Temporal Information Recognition from English Text Documents
Parul Patel, S. V. Patel · International Conference & Workshop on Emerging Trends in Technology · 2015
Temporal expressions are very important structure in a natural language. In order to use it in information retrieval, it needs to be extracted and normalized into its absolute value. In this paper we have presented a novel approach for temporal information extraction system from English documents. We have used CRF based classifier for extraction of temporal expression. System is evaluated it on Wiki war corpus and manually annotated documents. It achieves Precision of 91.3% , recall of 98.13% on data set 1 and 88.09% and 95.14% on data set 2 for detection of temporal expression.