As an essential dimension of our information space, time plays a very important role in every aspect of our lives. Temporal information is necessarily required in many applications, such as temporal constraint modeling in intelligent agents (Hritcu and Buraga, 2005), semantic web services (Pan and Hobbs, 2004), temporal content modeling and annotation for semantic video retrieval (QasemiZadeh et al., 2006), geographic information science (Agarwal, 2005), data integration of historical stock price databases (Zhu et al., 2004), ubiquitous and pervasive systems for modeling the time dimension of the context (Chen et al., 2004), and so on. Extracting temporal information from text is especially useful for increasing the temporal awareness for different natural language applications, such as question answering, information retrieval, and summarization. For example, in summarizing a story in terms of a timeline, a system may have to extract and chronologically order events in which a particular person participated. In answering a question as to a person’s current occupation, a system may have to selectively determine which of several occupations reported for that person is the most recently reported one (Mani et al., 2004). This chapter presents recent advances in applying machine learning and data mining approaches to automatically extract explicit and implicit temporal information from natural language text. The extracted temporal information includes, for example, events, temporal expressions, temporal relations, (vague) event durations, event anchoring, and event orderings.
Machine learning approaches have become more and more popular in extracting temporal information from text. This section focuses on the most recent efforts on extracting both explicit and implicit temporal information from natural language text.