Abstract: To promote the sharing positivity of big data alliance members from “inverted U” curve to “S” curve and improve the stability of alliance operations, the conception of “opportunistic sharing” strategy has been adoped to introduce alliance members. The strategy can not only enrich the traditional game strategy set, but also give more comprehensive reflects on the sharing behavior of big data alliance members under difference positive states. On this premise, this paper tries to construct a dynamic evolution game model for big data alliance, which has “no sharing-opportunistic sharing-sharing” strategy sets. It reveals the evolutionary game behavior of alliance members in positive and negative cases, and analyzes the key factors influencing the resource sharing positivity by using MATLAB to simulate the evolutionary path of the sharing game theory on big data alliance members. In the end, the results show that the social penalty coefficient has positive effects on the selection strategy of the alliance members, while the speculative yield coefficient negative. At the same time, the increase of the degree of sharing positivity is beneficial to the members of the alliance converting from the negative to the active sharing states.

Key words: big data alliance, data resource, evolutionary game, sharing positivity