Mechanical exploitation of the texts collected to supply modeling
Introduction to the notion
We introduce the hypothesis of the correspondence between the natural language categories and those of the object approach. On the evidence of this observation, textual analysis enables an almost mechanical passage from the texts collected towards the model. Several tools are based on this principle. The modeler can proceed manually, highlighting elements of text with different colors, according to their category: substantives (candidates for a class), relation verbs (associations), action verbs (operations), adjectives and participles (states), etc.
This technique is a good starting point for a modeler who feels a little lost when faced with the complexity or novelty of the domain studied. It encourages the domain descriptions to be collected or written.
The automatic or guided passage from text to model relies on the correspondence between the categories of natural language (grammatical categories) and those of modeling (categories of representation).