Analogy-based Text Normalization : the case of unknowns words (Normalisation de textes par analogie: le cas des mots inconnus) [in French]

Analogy-based Text Normalization : the case of unknowns words. In this paper, we describe and evaluate a system for improving the quality of noisy texts containing non-word errors. It is meant to be integrated into a full information extraction architecture, and aims at improving its results. For each word unknown to a reference lexicon which is neither a named entity nor a neologism, our system suggests one or several normalization candidates (any known word which has the same lemma as the spell-corrected form is a valid candidate). For this purpose, we use an analogybased approach for acquiring normalisation rules and use them in the same way as lexical spelling correction rules. Mots-clés : normalisation textuelle, correction orthographique, analogie.