The achievement of widespread immunity to infectious diseases like influenza by vaccination is very important for public health. Thus, it is essential to understand the drivers that contribute to voluntary vaccine uptake for the purpose of developing interventions to increase vaccine coverage. Previous studies have shown that there are multiple factors affecting an individual's intention to vaccinate, including safety of vaccines, severity of diseases, potential side effects and so on. However, limited studies have considered how individuals' perception, judgement and decision on voluntary vaccination may update due to the spread of social influence. In this work, an integrated model is proposed to analyse the dynamics of social influenced voluntary vaccination decision making, which takes into account both the formulation of individual's subjective judgements and the characterisation of influence spreading in a social network. It is assumed that individuals in social networks can explicitly and independently express their perceptions of vaccination. In the meantime, those who fail to make a firm decision initially can be influenced by their social neighbours. As a result, individuals' subjective judgements on vaccine uptake are characterised by belief distributions in the context of multiple criteria decision analysis (MCDA), whilst social influence is reflected in the form of belief spreading and updating against each of the decision criteria. The evidential reasoning (ER) rule, as a further development of the seminal DempsterShafer (D-S) theory of evidence, is then applied to aggregate multiple pieces of uncertain information resulted from both subjective judgements and social influence. A series of experimental simulations are conducted to analyse the dynamics of voluntary vaccination decision making under social influence in a social network. The effects of information flow and the sensitivity of individuals' preference towards certain criteria are explored in a systematic manner. This exploratory work provides a novel perspective to analyse social influenced decision making, and the research findings can be useful for disease control departments to better understand the dynamics embedded in an individual's vaccination decision as well as observe the trend of vaccinating behaviours in a social community.
