The invention relates to a method, used by computers, for developing an ontology (70) of a text (10) in natural language, comprising the steps of: receiving text data from the text (10); extracting the syntax and meaningful words (20) from the text via a grammatical analysis (S100) of the received data; for each of at least some of the meaningful words (20) of the text: searching (S200) for a definition (40) of that word (20) by means of at least one electronic dictionary (30, 35), extracting the syntax and the meaningful words of that definition, and creating (S300) an elementary lexical graph (50) of that definition based on the syntax and the meaningful words of that definition; and merging (S400) at least two of the elementary lexical graphs (50) created, as a function of the syntax of the text, so as to create at least one semantic graph (60) of the text. The invention further concerns a computer program product and an electronic data processing system comprising code means adapted to implement the steps of the method of the invention.