Implicative Statistical Analysis applied to clustering of terms taken from a pscyhological text corpus.

In order to validate a textual base contained in a behavioural skilltesting software, we suggest a methodology which can extract subsets of characteristic terms used to describe personality traits. Our approach permits, after an automatic language processing task, to evaluate the association rules between terms and descriptors (personality traits) structuring the corpus with the help of the theory of statistic implication.