Patent agents and researchers repeatedly face language and terminology barriers when checking the patentability of their own research results or when searching for suitable licensees. Existing patent retrieval methods are primarily based on textual searches in patent specifications and largely exclude illustrations and the references between text and image. Often, however, the innovation and exploitation potential of a patent can only be identified with the help of an illustration, and the analysis of a set of patents with similar or related innovations can be done by quickly looking at illustrations in a comparative way. Access via illustrations represents an alternative or supplement that functions independently of the language and terminology used and enables complete identification of exploitation-relevant results in patent specifications.
The project intervenes exactly at this point and develops a novel visual search for patent retrieval based on the automatic recognition of image similarities and text-image relations in patent specifications. The image-based search is based on machine learning techniques and is intended to increase the findability and visibility of patents and overcome language and terminology barriers. Measures from the areas of start-up promotion and technology marketing are used to open up new exploitation opportunities. Often, the innovation and exploitation potential of a patent can only be identified with the help of a mapping. This solution also enables domain and patent class independent patent searches and thus new ways of patent exploitation (keyword: cross-domain, cross-industry). The goal is to develop new methods for searching and analyzing visual elements in patents and to integrate them into the patent retrieval tool in order to meet the demand for innovation in the field of patent exploitation.