We present a system that couples techniques belonging to Information Extraction and deep linguistic processing for Question Answering. The system presented in the paper has undergone extensive testing and the parser has been trained on available testsuites and the Remedia corpus texts, one of which will be commented in some detail. We experimented our system with data made available within the QA4MRE task at CLEF 2011 and over 30 questions we got 24 correct answers. We realized that it is just thanks to the contribution of Situation Semantics and LFG based parsing that the system is able to select most appropriate answer candidates.
|Data di pubblicazione:||2011|
|Titolo:||Getask: A Hybrid System for Question Answering and Answer Recovery|
|Titolo del libro:||Proceedings of LERREW 2011|
|Appare nelle tipologie:||4.1 Articolo in Atti di convegno|
File in questo prodotto:
|getask.pdf||Abstract||Licenza non definita||Riservato|