Difference between revisions of "Temporal ordering"
| Line 3: | Line 3: | ||
For example, (taken from [[Yoshikawa_2009_jointly_identifying_temporal_relations_with_markov_logic|Yoshikawa et al. (2009)]]), consider the following sentence from a document: | For example, (taken from [[Yoshikawa_2009_jointly_identifying_temporal_relations_with_markov_logic|Yoshikawa et al. (2009)]]), consider the following sentence from a document: | ||
| − | :: With the introduction of the TimeBank corpus (Pustejovsky et al., 2003), machine learning approaches to temporal ordering became possible. | + | :: ''With the introduction of the TimeBank corpus (Pustejovsky et al., 2003), machine learning approaches to temporal ordering became possible.'' |
Here, temporal ordering task has to predict that the "Machine Learning becoming possible" event happened ''after'' the "introduction of the TimeBank Corpus" (event-event temporal relation) and ''overlap'' with the year 2003 (event-time temporal relation). Further, the task has to predict that both of these events occur ''before'' the document containing this sentence is created (event-document creation time temporal relation). | Here, temporal ordering task has to predict that the "Machine Learning becoming possible" event happened ''after'' the "introduction of the TimeBank Corpus" (event-event temporal relation) and ''overlap'' with the year 2003 (event-time temporal relation). Further, the task has to predict that both of these events occur ''before'' the document containing this sentence is created (event-document creation time temporal relation). | ||
Revision as of 16:32, 29 September 2011
This is a problem to automatically order events in text based on time. Temporal relation between any two events can be categorized into one of the 13 classes defined over TimeBank Corpus: before, ibefore (immediately before), includes, begins, ends, overlap and their inverses, plus simultaneous.
For example, (taken from Yoshikawa et al. (2009)), consider the following sentence from a document:
- With the introduction of the TimeBank corpus (Pustejovsky et al., 2003), machine learning approaches to temporal ordering became possible.
Here, temporal ordering task has to predict that the "Machine Learning becoming possible" event happened after the "introduction of the TimeBank Corpus" (event-event temporal relation) and overlap with the year 2003 (event-time temporal relation). Further, the task has to predict that both of these events occur before the document containing this sentence is created (event-document creation time temporal relation).
