graph embedding-based smart vaccination using mobile data

A novel smart vaccination method is proposed in this paper to distribute a limited number of vaccines among the people of a large community, such as a country, consisting of smaller communities like cities or provinces. the proposed method is comprised of two phases; a vaccine allocation phase and a targeted vaccination phase. in the first phase, the available vaccines are allocated to the communities based on demographics and the effectiveness of each type of vaccine. in the second phase, each community is modelled as a contact graph, and the vaccines available to the community are administered to the individuals whose vaccination has the greatest impact on breaking the chain of transmission. as a result of utilizing the node2vec graph embedding algorithm, the complexity of the proposed method increases linearly with the number of people in the community, as opposed to common centrality based methods, the complexities of which increase with the square or cube of the number of individuals. furthermore, the proposed method can distribute multiple types of vaccines with different probabilities of effectiveness. the performance of the proposed method is comparable to the common centrality based vaccination methods, while its complexity is lower. the results of the simulation show a 20% decrease in the peak number of infected individuals.