Dear all,
I work on a french banking corpus, I built a dedicated dictionary with a bank lexicon.
I have an ambiguity with the word "montant". He can be a Noun or a Verbe in present participle. In my corpus, he's never a Verb, but he was extracted like that with my graph.
(example from my corpus : "Cela correspond au montant à débloquer.")
So I defined priority rules on my dictionary (I followed the manual's procedure in the part 3.7.1), who contains :"montant,.N:ms".
I named this dictionary "dico_lexique_metier_banque-" with "-" at the end for assign the higher priority to the dictionary, and for tagging "montant" just as Noun, and not as Verb.
So with this procedure, after applying this dictionary and a graph who search the verbs, the result shouldn't contain "montant" as the present participle for the verb "monter".
But it doesn't work, my result is (where the value for "action" is the lemma for the verb): "{S}Cela correspond au <EtapePC Occurence="montant" Action="monter" Temps="Present"/> à débloquer."
Maybe I missed something or I didn't understand something ?
Does anyone know what the problem is, and can help me ?
For information, I use Unitex 3.1 beta under Windows 7.
Thanks in Advance
Gabrielle
Gabrielle