Resolución de anáforas que requieren conocimiento cultural con la herramienta FunGramKB

Autor: María de los Llanos Carrión Varela
Jazyk: English<br />Spanish; Castilian
Rok vydání: 2014
Předmět:
Zdroj: Revista de Lingüística y Lenguas Aplicadas, Vol 9, Iss 1, Pp 01-13 (2014)
Druh dokumentu: article
ISSN: 1886-2438
1886-6298
68184506
DOI: 10.4995/rlyla.2014.2003
Popis: While integrating linguistic knowledge of any kind is becoming an almost implicit practice in natural language understanding systems, the inclusion of cultural or world knowledge in these tools might have been neglected sometimes. However, a NLP system or knowledge base enriched with cultural information is a more robust, better cohesioned instrument for natural language understanding processes. The integration of this type of knowledge in NLP systems may be proven to contribute to solving some phenomena that occur in natural language, such as anaphor, metaphor and metonymy, ambiguity or co-reference, amongst others. The objective of this article is to describe the way FunGramKB (a knowledge base) integrates cultural knowledge in its conceptual modules and, in particular, how the information contained in the Onomasticon module of FunGramKB can contribute to maximising the informativeness and completeness of the whole system, thus resolving ambiguity problems in a determined linguistic phenomenon: anaphora.
Databáze: Directory of Open Access Journals