Voice-QA: Evaluating the Impact of Misrecognized Words on Passage Retrieval

Marcos Calvo (1) , Davide Buscaldi (2) , Paolo Rosso (1)

1 Departamento de Sistemas Informáticos y Computación [Valencia]
2 LIPN - Laboratoire d'Informatique de Paris-Nord

Question Answering is an Information Retrieval task where the query is posed using natural language and the expected result is a concise answer. Voice-activated Question Answering systems represent an interesting application, where the question is formulated by speech. In these systems, an Automatic Speech Recognition module can be used to transcribe the question. Thus, recognition errors may be introduced, producing a significant effect on the answer retrieval process. In this work we study the relationship between some features of misrecognized words and the retrieval results. The features considered are the redundancy of a word in the result set and its inverse document frequency calculated over the collection. The results show that the redundancy of a word may be an important clue on whether an error on it would deteriorate the retrieval results, at least if a closed model is used for speech recognition.

Mots clés: Passage Retrieval; Voice-activated QA

Domaines: Traitement du texte et du document

Marcos Calvo, Davide Buscaldi, Paolo Rosso. Voice-QA: Evaluating the Impact of Misrecognized Words on Passage Retrieval. 13th Ibero-American Conference on AI, Nov 2012, Cartagena de Indias, Colombia. pp.462-471. ⟨hal-00825246⟩