Publication

Abstract : Precise and timely information about opportunities for potential collaborations is very vital for the collaboration-intense research environment prevailing in innovation ecosystems. As the identification of suitable inventors for collaboration will be decisive for inventors in different phases of their careers, inventor collaboration recommendation systems are of great importance. Existing recommendation system frameworks for collaboration recommendations for academic authors and inventors are slightly intensive on the usage of link semantics. Like academic collaboration through co-authorship, collaborations of inventors through co-inventorship of patents can be found in almost all industrial areas in various degrees. Network representation of co-inventorship can be used to retrieve many insights that can even be vital for policymaking. In this work, for inventor collaboration recommendations, a minimal link semantics (MLS) approach based framework is built to overcome these major drawbacks and to improve usability. The case of inventors in the area ‘Wireless power transmission’ is analyzed using patent data for the demonstration of the MLS framework and on evaluation, the framework is found to be capable of retrieving novel and diverse recommendations to and from inventors that belong to different phases of a career.