Research Of Rumor Spreading Mechanism And Immunization Strategies

In recent years, many cases caused by rumor spreading occur frequently, suchas public panic, social disarray and severe economic loss, etc. Therefore, it is veryimportant to investigate the mechanism of rumor spreading, as well as to manipulatethe control of rumors effectively. Based on some realistic perspectives, this paperinvestigates the rumor propagation laws from several different aspects, and discussesthe prevention and control strategies for inhibiting the rumor spreading. The maininnovative achievements are listed as follows:1. A rumor spreading model considering remembering mechanism isestablished, and the decision-making evidence for dealing with rumor spreadingis presented under the mutual interaction of forgetting and rememberingThis dissertation introduces a new kind of people-Hibernator to reflect theinteraction effect of the forgetting and remembering mechanisms and supplementsone direct transition from ignorant to stifler. Then the classical rumor spreadingmodel has been extended and a novel rumor spreading model consideringremembering mechanism is developed, in which mean-field equations that describethe dynamics of the rumor spreading are derived. The steady-state analysis is carriedout to find out that the spreading threshold doesn’t exist in this model. Moreover, theemergence of forgetting and remembering mechanisms embodies the characteristicsof the repeatability on rumors, which makes the rumor termination time postponed.These require the government to race against time and to release the real informationrepeatedly to refute the rumor spreading.2. A double rumor spreading model considering two types of rumorsspread simultaneously is developed and the mutual interaction laws between thetwo rumors are discoveredConsidering the phenomenon that two rumors may appear during the sameperiod in the social network, a double rumor spreading model is built in an opensystem and the corresponding mean-field equations are also established. The basic reproduction number of the rumor spreading is obtained, and the local and globalasymptotically stability of the rumor-free equilibrium is proven. Moreover, the finalsize of the rumor is investigated. The simulation results show that there are bothcompetition and inhibition relationship between the two rumors.3. A rumor spreading model, involving people who see through the rumor,is built, and the optimal immunization proportion in homogeneous networksunder the random immunization strategy is calculatedThe immunized individuals are defined as a new group who see through therumor. The important role of the new group for the termination of rumor spreading isemphasized. In homogeneous networks, a rumor spreading model considering thepeople who see through the rumor is established. Moreover, following the analysisof the mean-field equations describing the dynamics of the modified rumorspreading process, the spreading threshold and the immunization threshold under therandom immunization strategy are calculated. The government implementsimmunization to a certain percentage of the population to prevent the spread ofrumors. These findings suggest that immunizing a higher proportion of individuals isnot necessarily better due to the social costs and resource wastes. The optimalsolution is to choose an immunization rate equal to the immunization threshold.4. A rumor spreading model considering forgetting mechanism in complexnetworks is established, and the effects of different immunization strategies forprevention and control rumor spreading are examinedThis dissertation develops a rumor spreading model considering forgettingmechanism in complex networks. The mean-field equations are derived to describethe dynamics of the rumor spreading in complex networks. The steady-state analysisis carried out, indicating that the spreading threshold exists in both homogeneousand inhomogeneous networks. Afterwards, immunization strategies are implementedto hinder rumor spreading. In homogeneous networks, the immunization thresholdunder the random immunization strategy is obtained. In inhomogeneous networks the spreading thresholds under the random immunization and targeted immunizationstrategies are obtained, respectively. Numerical simulations reveal that the networktopology exerts significant influence on the rumor spreading process, and that thetargeted immunization works better in the BA scale-free network for rumors controlthan the random immunization.