Innovation Capability (IC) are important strategic asset used by companies for providing and sustain their competitive advantage. IC is the firm's ability to mobilize and create new knowledge applying appropriate process technologies. The main purpose of the paper is to understand the patent's features that are relevant to predict IC. Starting from patent data, patent's forward citations are used as proxy of IC and the main patents features are considered as proxy of IC determinants. For this analysis, patents issued by firms with registered office in Italy and Sweden or by inventors with residence in Italy or Sweden are investigated, using three different algorithms of machine learning: Least Squares (RLS), Deep Neural Networks (DNN), and Decision Trees (DT). Results are two-fold. First, from a methodological perspective, machine learning approaches are useful in reducing the number of features needed to explain IC, and so in reducing the complexity of the analyses by focusing only on the features with high predictive power. Second, from a managerial standpoint, the study suggests which few, but relevant, variables managers should look at in writing and issuing patents.
Patents and big data to forecast firms' innovation capability
Ponta, Linda;Oneto, Luca;
2019-01-01
Abstract
Innovation Capability (IC) are important strategic asset used by companies for providing and sustain their competitive advantage. IC is the firm's ability to mobilize and create new knowledge applying appropriate process technologies. The main purpose of the paper is to understand the patent's features that are relevant to predict IC. Starting from patent data, patent's forward citations are used as proxy of IC and the main patents features are considered as proxy of IC determinants. For this analysis, patents issued by firms with registered office in Italy and Sweden or by inventors with residence in Italy or Sweden are investigated, using three different algorithms of machine learning: Least Squares (RLS), Deep Neural Networks (DNN), and Decision Trees (DT). Results are two-fold. First, from a methodological perspective, machine learning approaches are useful in reducing the number of features needed to explain IC, and so in reducing the complexity of the analyses by focusing only on the features with high predictive power. Second, from a managerial standpoint, the study suggests which few, but relevant, variables managers should look at in writing and issuing patents.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.