Abstract: Through feature analysis and algorithm analysis, transmission evolution of on-line trading evaluation during consumption is studied In consideration of the on-line customers’ viewpoint inclination, combining the cluster attribute of on-line trading evaluation viewpoints, the conversion rule of consumers’ viewpoint inclination, and as well as the e-business trading evaluation evolution and migration cell model which is based on small world network effect are put forwarded, and the evolution process of evaluation viewpoints is analyzed Through the experiment of changing the clustering and classification design of beverage goods on a certain trading platform, customers’ consumption psychology and shopping orientation can be forecasted, which will help cultivate customers’ loyalty, and increase the sales volume

Key words: complex network, on-line trading, opinion transmission, on-line evaluation