The principal aim of the MODELACT project is to pass from the identification of the action concepts produced by the IMAGACT project to their definition, supplying valid modelling especially for natural language processing and man-machine interaction.
The project will lead to crucial advancements in our knowledge in the field of human cognition, and will specify which are cognitively the most important qualities for segmentation of the continuum of the action space in discrete units encoded by natural language.
This ambitious objective is now realistic since it is based on a wide set of data.
In this respect, MODELACT constitutes a change in the semantic paradigm, from deductive to inductive, in the creation of ontologies.
The problem of the type of action models processed by the human cognitive system is faced from different points of view, that will be integrated during the development of the project: acquisitional, semantic and computational.