The community structure in fully signed networks that considers both node attributes and edge signs is important in computational social science; however, its physical description still requires further exploration, and the corresponding measurement remains lacking. In this paper, we present a generalized framework of community structure in fully signed networks, based on which a variant of modularity is designed. An optimization algorithm that maximizes modularity to detect potential communities is also proposed. Experiments show that the proposed method can efficiently optimize the objective function and perform effective community detection.
Missing data caused by boundary specification has a detrimental effect on the analysis of network structures, and designing optimal sampling methods is crucial for conducting network investigations. The present study discusses the boundary specification problem in multiple surveys, and proposes a mathematical model for optimizing the sampling strategy in each independent survey. A memetic algorithm that maximizes the sample representativeness is proposed as well, and experiments have proved the effectiveness and efficiency of the proposed algorithm. Zachary’s Karate Club network and three networks of migrant workers are also performed to explain the social meaning of the optimal sampling method.
Small-world effect plays an important role in the field of network science, and optimizing the small-world property has been a focus, which has many applications in computational social science. In the present study, we model the problem of optimizing small-world property as a multiobjective optimization, where the average clustering coefficient and average path length are optimized separately and simultaneously. A novel method for optimizing small-world property is then proposed based on the multiobjective evolutionary algorithm with decomposition. Experimental results have proved that the presented method is capable of solving this problem efficiently, where a uniform distribution of solutions on the Pareto-optional front can be generated. The optimization results are further discussed to find specific paths for optimizing different objective functions. In general, adding edges within the same community is helpful for promoting ACC, while adding edges between different communities is beneficial for reducing APL. The optimization on networks with the feature of community structure is more remarkable, but community structure has less impact on the optimization when the internal community is triangles-saturated.
This chapter draws on data from a 2020 survey conducted in Shaanxi Province during the COVID-19 outbreak to investigate the relationship between family resources and children’s psychological well-being among four main groups of children in China: urban children, migrant children, rural non-migrant children, and rural left-behind children. The analysis reveals the multifaceted ways in which the pandemic has interacted with family resources to shape children’s well-being. Although all groups experienced a decline in household economic resources, left-behind children were disproportionately affected by the economic shock. At the same time, parent-child relationships improved across the board, suggesting a potential compensatory effect. Reduced economic resources emerged as a risk factor, whereas enhanced familial relationships served as a protective factor for children’s psychological outcomes. Notably, for migrant and left-behind children—groups typically deprived of parental involvement—the strengthening of parent-child bonds had a particularly large positive effect on psychological outcomes, thereby narrowing the well-being gap between them and their urban peers during the pandemic. In contrast, the influence of family economic resources was less pronounced and did not exhibit clear variation across the groups. Overall, these findings provide new insights into how large-scale crises like the COVID-19 pandemic interact with family resources to shape child well-being in contemporary China.
Data obtained from experiments and surveys in human social systems are inevitably influenced by systematic measurement errors, and network data are no exception. Despite the prevalence of error in social network data, current research often lacks rigorous estimation of its expected precision, which may lead to biased conclusions. Signed networks, which encode both positive and negative relationships, constitute an important component of network science, and conducting measurement error analysis on them can substantially enhance the accuracy of social network analysis. This paper proposes a set of error measurement tools based on the Expectation-Maximization (EM) algorithm, specifically designed to estimate errors in signed network data. We extend traditional experimental error estimation to the network domain, derive a general error estimation method for signed networks, and validate its scientific validity and practical utility through extensive simulation experiments on both synthetic and real-world networks. The experiments reveal that network density and the ratio of positive to negative edges significantly influence the posterior probability distribution of the adjacency matrix. Specifically, as density increases, edge estimation accuracy exhibits a U-shaped trend, and the proportion of negative edges shows a nonlinear relationship with accuracy. The proposed method is applicable to repeatedly measured signed networks and provides a reliable framework for reconstructing network structures as faithfully as possible.