| dc.description.abstract | Innovations are fundamental for a sustainable growth of science. Studying the process of innovation and understanding how rate of discovering creative and new ideas can be optimized is essential in order to ensure the continued propagation of novelty in science. However, the lack of a foolproof method of finding novel concepts makes finding efficient innovation research strategies difficult. Past solutions in the form of innovation models have been able to simplify aspects of this process and achieved some success related to the ideation of optimal strategies but have failed to address key aspects of research such as innovation communication, heterogeneity of researchers and effect of failures, among other things.
This thesis focuses on generating new tools for designing strategies of finding innovations. It uses bio-inspiration in the form of ant foraging by building a comprehensive analogy with the innovation process. A simulation model is constructed that uses a bibliometric network built from the metadata of articles within a particular scientific field as the environment, and specifically configured ants as agents interacting with the environment. Multiple aspects of ant foraging that improve efficiency of finding food were shown to have a direct correlation with research aspects that, if followed, can boost innovative discoveries. Some of these aspects include efficient communication of innovation, publishing failures and inconclusive research, and a balance between conservative and risky research.
Finally, the simulations showed that configuring the parameters so that the agents behaved exactly as some of the exemplary species of ants, such as Sahara Desert ants, Argentine ants, etc., results in highly efficient foraging. These parameters correspond to strategies that the ants employ during the foraging process, and the strategies translate to actions that researchers can take to improve the innovation finding rate. Thus, the thesis generates new tools for improved novelty discovery and addresses the limitations of other innovation models. | |
