Natural Language Processing in assistance to Inventive Design activities

Design activity requires engineers to use their knowledge understand, formulate and solve industrial problems whose complexity is constantly increasing. Our work, at the frontier of linguistics, computer science engineering sciences, consists in making optimal information contained texts useful designers, particularly patents. Existing method extract from patents already exist, but without taking into consideration output as a mean populate Inventive ontology. This paper describes how we Natural Language Processing designers context an inventive activity. ontology inspired by Theory Problem Solving (TRIZ) helps us categorize collected corpus consisting thousands unstructured texts, redistribute it designer thus increase his ability that subject study confronts him with. concerns new way categorizing newly defined technical domains so better perceived prior art during early stages process. results attest significant progresses both terms speed understand data accuracy perception explored field enter resolution logic.