Skip to Main content Skip to Navigation
Master Thesis

Recherche automatique d'antériorités de brevets par la recherche de revendications et de segments proches potentiellement invalidants

Abstract : Our internship took place in Tecknowmetrix; a business specialized in innovation, intellectual property and strategic intelligence. Our aim here was to create a tool which allows validating or invalidating some patent claims, in an automatic or semi-automatic way. Our goal was to invent a priority research tool which finds all the relevant and similar patents in order to find segments which are invalidating "novel criteria". Our methodology is different from classical ways which are exploiting statistical methods and is more oriented towards linguistic in basing itself on the internal structure of patents. Our work involved in natural language processing field. To do so, we defined what a patent is and we studied claims by lexical and syntax views in order to discover some regularities. These regularities will help us to figure out how producing a data model which can be applied on our tool that we want to develop. Our tool should be able to extract similar segments of similar claims on a corpus of patents written in the English language.
Document type :
Master Thesis
Complete list of metadata

Cited literature [8 references]  Display  Hide  Download

https://dumas.ccsd.cnrs.fr/dumas-00631733
Contributor : Uga - Bulles <>
Submitted on : Monday, September 3, 2012 - 2:42:47 PM
Last modification on : Sunday, March 14, 2021 - 9:54:03 PM
Long-term archiving on: : Saturday, January 14, 2012 - 2:26:27 AM

File

 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2031-01-18

Please log in to resquest access to the document

Identifiers

  • HAL Id : dumas-00631733, version 1

Citation

Edmée Marazel. Recherche automatique d'antériorités de brevets par la recherche de revendications et de segments proches potentiellement invalidants. Linguistique. 2011. ⟨dumas-00631733⟩

Share

Metrics

Record views

444