An Agent-Based Innovation Diffusion Model In The Small World Network

The improvement of technology and the development of the human society are producing more innovations nowdays. More and more enterprises and countries put the innovation as the most important strategy, which can be found in the phenomena many industry clusters and innovation centers are built. The purpose is to create more innovations and let them diffuse in the network effectively.In such world circumstance, we can see more and more innovation diffusion models research. From the very beginning we used the math model to simulate the innovation diffusion and now we use CA models and Agent-based models. To name just a few. And this paper design an agent-based innovation diffusion model which has the feature of 10-steps memory on the basis of small world network and the model is constructed in matlab. Then we simulate the one-innovation and two competitive innovation diffusion, which is used to analyse the factors which will affect the diffusion and conclude the optimal strategies.By the simulation, we find in one-innovation diffusion, the scatter diffusion is better than the cluster diffusion. And increase the'Innovation Sensitiveness' will decrease the possibility of'Lock-In'. So in the real world, we can use the strategy like advertisement to increase the 'Innovation Sensitiveness', which will make the diffusion efficient. And in the two competitive innovation diffusion, we find when the competitive policy works, more effective the word-of-mouth strategy will be. And when the competitive policy doesn't work, more effective the advertisement strategy will be. So the innovation in the real world should launch the suitable strategy to get the optimal diffusion.