We propose a simple probabilistic model of zapping behavior of televisionviewers. Our model might be regarded as a `theoretical platform' to investigatethe human collective behavior in the macroscopic scale through the zappingaction of each viewer at the microscopic level. The stochastic process ofaudience measurements as macroscopic quantities such as television programrating point or the so-called gross rating point (GRP for short) arereconstructed using the microscopic modeling of each viewer's decision making.Assuming that each viewer decides the television station to watch by means ofthree factors, namely, physical constraints on television controllers,exogenous information such as advertisement of program by television station,and endogenous information given by `word-of-mouth communication' through thepast market history, we shall construct an aggregation probability ofGibbs-Boltzmann-type with the energy function. We discuss the possibility forthe ingredients of the model system to exhibit the collective behavior due tonot exogenous but endogenous information.
展开▼
