Abstract: In this paper, the formation of inter-organization innovation network during the process of knowledge integration is modeled. By simulation analysis, it is found that the complementary extent of inter-organization and the efficiency of knowledge integration have a certain impact on the structure of innovation network. More specifically, the innovation network has the property of random network, i.e., short average path length and low clustering coefficient, when the complementary extent of inter-organization and the efficiency of knowledge integration are high; the innovation network has the property of regular network, i.e., long average path length and high clustering coefficient, when the complementary extent of inter- organization and the efficiency of knowledge integration are low; the innovation network has the property of small-world network, i.e., relatively short average path length and high clustering coefficient, when the complementary extent of inter- organization and the efficiency of knowledge integration are neither too high nor too low.

Key words: knowledge integration, innovation network, small-world network, clustering coefficient, average path length