Characterization of Innovative Technologies in Healthcare 4.0 Through the Analysis of Italian Patents

Identifying promising and frontier technologies is a fundamental step for technological management, especially for planning R&D policies by governments and companies. Among the intellectual property (IP) assets, the patents have a strategic role because they contain from 70 to 90% of information about technologies. One of the most prominent phases in managing patents is their classification according to a proposed taxonomy, a particularly expensive and time-consuming phase, above all due to the increase in the number of filed patents, the complexity of the contents and the difficulty in classify them in the right category. Users that search for a potentially useful patent, often deal with “monster class”, multidisciplinary categories containing a significant number of poorly characterized and classified patents. Monster categories are thus ineffective, not discriminating, and difficult to explore. A typical example of “monster class” is often present in the healthcare related patents, due to its multidisciplinary and particularly innovative nature. The paper proposes an approach for: i) improving taxonomy-based classification by using Natural Language Processing and clustering techniques; ii) improving keywords-based classifications by using a complex network analysis on the most frequent words. The approach was experimented on Knowledge-Share platform (an Italian patent platform) with a particular attention on patent related to healthcare 4.0 with the aim of identifying the technology trends in this cross-domain context, where artificial intelligence is identified to be having an impact, such as in robotic assisted surgery, personalized medicine, drug discovery, imaging & diagnostic, diseases monitoring, assisted living, automated clinical decision support, elderly care, and so on.