Information-Driven Collective Intelligences

Information-driven collective intelligences derive from the connection and the interaction of multiple, distributed, independent agents that produce and process information, and eventually turn it into knowledge, meant in the broad sense of capability of conceptual representation. For instance, the Worldwide Web can be viewed as an information-driven collective intelligence emerging from the digital network known as the Internet. The point here is how to extend such a capability for knowledge generation from the participating agents to the collective intelligence itself. In this paper we show how this can be obtained with graph-based algorithms for the detection of communities of agents so as to support a dynamic, self-organized form of concept-discovery and concept-incarnation. In particular, we show how to strengthen community ties around concepts in order to increase their level of socialization and, consequently, of "fertility" in the generation of new concepts. Since there exists a direct relationship between concept discovery and innovation in human intelligences, we point out how analogous innovation capabilities can now be supported within information-driven collective intelligences, with direct applications to product innovation.