Formation and recurrence mechanism of residents' waste separation behavior under the intervention of an information interaction

Abstract In the context of socialization, policy guidance and psychological bias play key roles in the recurrence of urban household waste separation behavior. This study explores the recurrence path of urban residents' waste separation behavior from the perspective of behavioral reinforcement, and constructs a reinforcement learning model of the separation behavior, with an interactive intervention of contextual information and biased information in the situation of policy response control. On this basis, this study clarified the realistic conditions and optimal path of urban residents' waste separation behavior recurrence under an interactive intervention involving multiple sources of information through computer simulation. The study found that, in the situation of fixed response validity, the effect of the interactive intervention can be strong, weak or negative, which correspond to optimal, sub-optimal and negative information intervention sets. In the situation of different degrees of response validity, a high degree of response validity leads to the fastest formation and recurrence of separation behavior. High degree of understanding and low degree of support lead to only a high individual learning rate in the initial stage, while low degree of understanding and high degree of support lead to a higher learning rate in the following period time.