Selecting answers with structured lexical expansion and discourse relations

Martin Gleize (1) , Brigitte Grau (2, 1) , Anne-Laure Ligozat (1, 2) , Van-Minh Pho (1) , Gabriel Illouz (1) , Frédéric Giannetti (1) , Loic Lahondes (1)

1 LIMSI - Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur
2 ENSIIE - Ecole Nationale Supérieure d'Informatique pour l'Industrie et l'Entreprise

n this paper, we present the LIMSI’s participation to QA4MRE2013. We decided to test two kinds of methods. The first one focuses on complex questions, such as causal questions, and exploits discourse relations. Relation recognition shows promising results, however it has to be improved to have an impact on answer selection. The second method is based on semantic variations. We explored the English Wiktionary to find reformulations of words in the definitions, and used these reformulations to index the documents and select passages in the Entrance exams task.

Mots clés: question answering; index expansion; discourse relation; question classification

Domaines: Informatique [cs]; Informatique et langage [cs.CL]

Martin Gleize, Brigitte Grau, Anne-Laure Ligozat, Van-Minh Pho, Gabriel Illouz, et al.. Selecting answers with structured lexical expansion and discourse relations: LIMSI's participation at QA4MRE 2013. QA4MRE Workshop, CLEF conference, Sep 2013, Valencia, Spain. ⟨hal-02296905⟩