Section: New Results

PerLex, a morphological lexicon for the Persian language

Participant : Benoît Sagot.

In the context of the PerGram project 8.2.2 , Benoît Sagot collaborated with Géraldine Walther (LLF, Université Paris 7) and Pollet Samvélian (Université Paris 3) to begin the development of a morphological and syntactic lexicon for the Persian language, as well as a processing chain (i.e., a Persian version of Sx Pipe). In 2009, the first step towards this goal has been achieved, and the first version of the PerLex lexicon has been released [38] . It only contains morphological information (valency frames and complex predicates are planned for 2010), has not been manually validated yet apart from verbs and some specific entries (a full validation by native speakers is planned in the first half of 2010), and is still to be completed and augmented thanks to techniques described in 6.4 . But this is the first large-coverage freely available lexicon for Persian.