| dc.description.abstract | Many universities have as strategic goal to raise their standards of innovation and Research and Development (R&D). One of the main measures adopted is to promote the externality of intellectual property through the filing of patents. Recent research proposes that working collaboratively with industry, government and other stakeholders helps to develop innovation and is essential to ensure the long-term and prestigious academic institutions. However, we do not observe in the literature an emphasis on studies that present the tactical and strategic paths for science and technology institutions to potentiate the generation of patents. Given this scenario, this research aims to identify which combinations of the relational structure of the university with external actors best influence the results of R&D. To identify these combinations, the present research used two research methods: Social Network Analysis (SNA) and Qualitative Comparative Analysis (QCA). The data used in this study were collected from the Derwent Innovation Index (DII) database with more than 16 million inventions registered from the world's leading patent offices. With the help of the Gephi software, the network data related to the global semiconductor industry were analyzed. Subsequently, with the help of the fsQCA software, the combinations (recipe) responsible for the success cases in the generation of patents were generated in the three periods of analysis between 1963 and 2017. As a result of the research, one can in the first period of analysis (1963-1990), considered a period with little competitiveness in the semiconductor industry, the direct and recurrent relationship between the actors was the main propeller of generation of patents. In the second period of analysis (1991-2001), where the semiconductor industry began to move towards more participatory R&D projects, the combination that generated the best results had the configuration of actors with direct and recurrent relationships, acting in a close fashion - few ways to access the other actors of the network - and organizing in clusters, that is, acting with high level of grouping. The third period (2002-2017), characterized by a highly competitive and dynamic market in this type of industry, the two recipes that generated the best results or combined direct and recurrent relationships with actors from a few distant ones, or again combined direct and recurrent relations with actors who in turn have direct contacts with many actors, that is, actors who have a lot of influence and prestige when compared to others. Finally, research contributes to the theory by filling a gap in the study of the most effective relational structures in the generation of intellectual property. For the managerial field, the study indicates that to maximize the generation of patents in universities, it is necessary to position the university in a strategic way with close relations, recurrent with actors with a high degree of influence and prestige. | en |
