Methods for improving efficiency of morphosyntactical analyzer for French information processing

It is well known that the efficiency of morpho-syntactic analysis depends, among others, on the structure of the dictionary, the algorithm and linguistic philosophy adopted to classify word types and analyze suffixes of irregular verbs. This paper proposes rules to make more efficient the morpho-syntactic analysis, in particular, of French. To do this, the authors try first to classify the forms of French words according to their compositional characteristics. They then propose a series of rules breaking these words into smaller words or morphemes. Additional decomposition rules to analyze complex sentences into simple ones are also presented. With these decomposition rules, an improvement on the morpho-syntactic analyzer for French can be expected