| Autor: |
Sonbol, Riad, Rebdawi, Ghaida, Ghneim, Nada |
| Předmět: |
|
| Zdroj: |
International Journal of Electrical & Computer Engineering (2088-8708); Jun2018, Vol. 8 Issue 3, p1766-1773, 8p |
| Abstrakt: |
Anaphora resolution (AR) is one of the most important tasks in natural language processing which focuses on the problem of resolving what a pronoun, or a noun phrase refers to. Moreover, AR plays an essential role when dealing with business process textual description, either when trying to discover the process model from the text, or when validating an existing model. It helps these systems in discovering the core components in any process model (actors and objects).In this paper, we propose a domain specific AR system. The approach starts by automatically generating the concept map of the text, then the system uses this map to resolve references using the syntactic and semantic relations in the concept map. The approach outperforms the state-of-the art performance in the domain of business process texts with more than 73% accuracy. In addition, this approach could be easily adopted to resolve references in other domains. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
