Automatic temporal event recognition from Thai news

In this paper, we propose a technique about Automatic temporal extraction of chronological information in natural language processing (NLP). Temporal information extraction has value in question-answering, human decision-making and summarization. Time-revealing vocabularies are investigated and grouped into a lexical pattern of temporal expression. Events and temporal expression time were extracted and temporal expression time transform into calendar time. Then, simple calendar time is used to order temporal time and their events. To convert time expression into absolute time we used a rule-based approach. The effectiveness of our approach was evaluated and the correctness was around 68%.