Risk perception plays a crucial role in shaping health-related behaviors in a variety of infectious disease control settings. The purpose of this study was to assess risk perception and behavioral changes in response to influenza epidemics. We present a risk perception assessment model that uses information theory linking with a probabilistic risk model to investigate the interplay between risk perception spread and health behavioral changes for an influenza outbreak. Building on human influenza data, we predicted risk perception spread as the amount of risk information. A negative feedback-based information model was used to explore whether health behavioral changes can increase the control effectiveness. Finally, a probabilistic risk assessment framework was used to predict influenza infection risk based on maximal information-derived risk perception. We found that (i) an individual who perceived more accurate knowledge of influenza can substantially increase the amount of mutual risk perception information, (ii) an intervening network over which individuals communicate overlap can be more effective in risk perception transfer, (iii) collective individual responses can increase risk perception information transferred, but may be limited by contact numbers of infectious individuals, and (iv) higher mutual risk perception information gains lower infection risk probability. We also revealed that when people increased information about the benefits of vaccination and antiviral drug used, future infections could significantly be prevented. We suggest that increasing mutual risk perception information through a negative feedback mechanism plays an important role in adaptation and mitigation behavior and policy support.