Difference between revisions of "Narrative event chains"
| Line 1: | Line 1: | ||
| − | This is a [[category::problem]] to | + | This is a [[category::problem]] to temporally order sets of narrative events centered around a common protagonist. A narrative event is a tuple of event (a verb) and its participants' typed dependencies. An example of a pair of narrative events related by a common protagonist is ''e(push,patient)'' and ''e(fall,agent)'' in which the patient of ''push'' is the agent of ''fall''. |
Narrative chains are related to structured sequences of participants and events that are called scripts ([[RelatedPaper::Roger C. Schank and Robert P. Abelson. 1977. Scripts, plans, goals and understanding. Lawrence Erlbaum|Schank and Abelson, 1977]]). For example, (taken from [[Chambers_and_Jurafsky,_Unsupervised_Learning_of_Narrative_Event_Chains,_ACL_2008|Chambers and Jurafsky (2008)]]), an example of a narrative chain: | Narrative chains are related to structured sequences of participants and events that are called scripts ([[RelatedPaper::Roger C. Schank and Robert P. Abelson. 1977. Scripts, plans, goals and understanding. Lawrence Erlbaum|Schank and Abelson, 1977]]). For example, (taken from [[Chambers_and_Jurafsky,_Unsupervised_Learning_of_Narrative_Event_Chains,_ACL_2008|Chambers and Jurafsky (2008)]]), an example of a narrative chain: | ||
Revision as of 01:03, 29 November 2011
This is a problem to temporally order sets of narrative events centered around a common protagonist. A narrative event is a tuple of event (a verb) and its participants' typed dependencies. An example of a pair of narrative events related by a common protagonist is e(push,patient) and e(fall,agent) in which the patient of push is the agent of fall.
Narrative chains are related to structured sequences of participants and events that are called scripts (Schank and Abelson, 1977). For example, (taken from Chambers and Jurafsky (2008)), an example of a narrative chain:
_ accused X
X claimed _
X argued
_ dismissed X
Another example of narrative chain is:
W joined _
W served _
W oversaw _
W resigned
The problem of automatic extraction of narrative event chains can be broken down into three separate sub problems: (1) extraction of events sharing co-referring arguments, (2) temporal ordering of extracted events, (3) pruning and clustering of self-contained chains from the space of events.
