ExpResViP – Verwertung von Forschungsergebnissen durch visuelles Patentretrieval
Unlike purely text-based search methods, visual patent retrieval is based on the automatic recognition of image similarities and text-image relationships. This allows the range of searches to be extended across patent classes and domains, enabling, for example, the identification of new markets or licensees outside of the known areas of application. Furthermore, language-independent visual searches can be used in foreign-language databases if no adequate full-text translations are available.
The BMBF-funded project is being carried out from August 2020 to July 2023 in cooperation with the TIB – Leibniz Information Centre for Science and Technology and the Fraunhofer Institute for Intelligent Analysis and Information Systems (IAIS). The role of Leibniz Headquarters is to develop skills in visual patent retrieval and to prepare for use of the new tool, for example in the Start-up Advisory Service.
In order to ensure that the new research tool is efficient and sustainable to use, its development is being driven not purely by technology, but from a user-oriented perspective. For this purpose, the needs and user behaviour are analysed at regular intervals and the tool is iteratively adapted based on the feedback from these analyses.
Interested Leibniz Institutes are therefore warmly invited to take part in the project, to test the prototypes of the new tool and to support its development by providing feedback.