Abstract:

The method of detecting functional knowledge community and function innovation from the international patent classification (IPC) is proposed. Through the extraction and analysis of the co-occurrence relationship of functional verbs from IPC, a functional verb co-occurrence network is constructed. The CNM algorithm is used to detect communities in the functional verb co-occurrence network. Four typical topological relations between the nodes of the network are summarized. The topological relations between the internal and inter community nodes are analyzed. According to the topological features of the node connection of the functional verbs, four types of functional intermediary nodes are defined, and four strategies are proposed to realize product function innovation. As an application example, functional verbs from all entries of IPC with co-occurrence relationship with draining and drying are extracted, and a functional verb co-occurrence network including 85 nodes is constructed. A pipeline lighting system with function of regulating brightness using power supply from drainage energy is invented successfully.

Key words: community detection and reorganization, co-occurrence network, function innovation, functional verb, international patent classification (IPC)