Revue: | Polibits |
Base de datos: | PERIÓDICA |
Número de sistema: | 000368142 |
ISSN: | 1870-9044 |
Autores: | Murata, Masaki1 Utiyama, Masao1 Kanamaru, Toshiyuki1 Isahara, Hitoshi1 |
Instituciones: | 1National Institute of Information and Communications Technology, Kyoto. Japón |
Año: | 2009 |
Periodo: | Jul-Dic |
Número: | 40 |
País: | México |
Idioma: | Inglés |
Tipo de documento: | Artículo |
Enfoque: | Experimental, aplicado |
Resumen en inglés | We describe a method of using multiple documents with decreasing weights as evidence to improve the performance of a question–answering system. We also describe how it was used in cross–language question answering (CLQA) tasks. Sometimes, the answer to a question may be found in multiple documents. In such cases, using multiple documents for prediction generates better answers than using a single document. Therefore, our method uses information from multiple documents by adding the scores of candidate answers extracted from the various documents. Because simply adding scores degrades the performance of question–answering systems, we add scores with decreasing weights to reduce the negative effect of simply adding. We used this method in the CLQA part of NTCIR–5. It was incorporated into a commercially available translation system that carries out cross–language question–answering tasks. Our method obtained relatively good CLQA results |
Disciplinas: | Ciencias de la computación |
Palabras clave: | Procesamiento de datos, Aprendizaje de máquinas, Traducción, Documentos múltiples, Lingüística computacional |
Keyword: | Computer science, Data processing, Machine learning, Translation, Multiple documents, Computing linguistics |
Texte intégral: | Texto completo (Ver HTML) |