TERMIANE is a project started at CNRS-LIPN in 1997 and co-developed with CNRS-IRIT since 1998. The TERMINAE method and tool aim at guiding knowledge acquisition from texts to build terminological and ontological models. The target conceptual has the particularity to combine a lexical and a conceptual component. The method suggests a supervised modeling process. Text analysis is supported by several NLP tools (such as term extractor and relation extractor with lexico-semantic patterns) whereas the TERMINAE modeling workbench keeps track of the various status of the knowledge represented in the model, from text fragments to lexical structures (terminological forms) and then conceptual and formal structures. The method is divided into 4 stages, which may be more or less important according to the kind of application in which the model will be used: corpus selection and organisation; linguistic analysis with the help of several NLP tools; normalization according some structuring principles and criteria; formalization and validation with a variant language of description logics. TERMINAE has been evaluated through several case studies and industrial cooperation to build terminologies, application specific domain models or corporate ontologies.

For more information on this topic see also the TERMINAE-website.
contact initiatives about topics events players