The ILIAD Project : Analysing Information using InformetricsTechniques and Natural Language

We present the ILIAD project and its current results. ILIAD aims at combining statistic and linguistic approaches in order to analyse information in large documentary databases. The resulting analysis should enable a human operator to collect the information content of a set of texts without having to read it sequentially. Our current experimentation concerns the analysis of a set of abstracts extracted from a documentary database. Our approach relies upon the recent terminology advance in both linguistics and knowledge aspects. In a rst step we identify terms in texts and classify them using a statistical algorithm. The second step is a partial linguistic analysis which focusses on terms highlighted by the classiication process. Finally, techniques form artiicial intelligence are called upon in order to collect and organise the information that emerges from these texts.