A new question about modeling information

I read a graduate economics textbook by Jehle and Reny tonight. The generalization of utility theory based on preference relations makes me think about such a question:

Can interactive information modeling be generalized similarly. Savage has done a job of using preference relations to understand beliefs. But there is no interation among decision makers.

Is belief similar to utility in the sense that it can be generalized? what can we learn from such generalization? what can be the role of multidimentionality (awareness) of information in this research?

So here are many abstract questions to address. It may be a nice topic to study for a long term in the future, if there is a chance.