Social System Inference from Noisy Observations

Abstract

This paper studies social system inference from a single trajectory of public evolving opinions, wherein observation noise leads to the statistical dependence of samples on time and coordinates. We first propose a cyber-social system that comprises individuals in a social network and a set of information sources in a cyber layer, whose opinion dynamics explicitly takes confirmation bias, negativity bias and process noise into account. Based on the proposed social model, we then study the sample complexity of least-square auto-regressive (AR(1)) model estimation, which governs the number of observations that are sufficient for the identified model to achieve the prescribed levels of accuracy and confidence. Building on the identified social model, we then investigate social inference, with particular focus on the weighted network topology, the subconscious bias and the model parameters of confirmation bias and negativity bias. Finally, the theoretical results and the effectiveness of the proposed social systems are validated by the US Senate Member Ideology data.

Cost Function Learning in Memorized Social Networks with Cognitive Behavioral Asymmetry

“…A similar version building on Friedkin-Johnsen model [27] is proposed in [9], [41], [42] to capture confirmation bias, since both polarization and homogeneity are the results of the conjugate effect of confirmation bias and social influence [34], [38]. However, the model can only capture symmetric confirmation bias.…”

Dynamics of Public Opinion Evolution with Asymmetric Cognitive Bias

“…For the rest of proof, we consider contradiction, i.e, assuming (11a)-(11d) do not hold, the condition (7a) does not hold as well. We assume that (11) does not hold, i.e., g i (f i (x i ) − f i (x j )) is nondecreasing w.r.t.|f i (x i ) − f i (x j )|. We let x i ≥ x j > 0, and thus have…”

Recovering unknown topology in a two-layer multiplex network: One layer infers the other layer

“…However, in reality, network topologies are often completely or partially unknown, e.g., in the situations of human gene expression networks, biological neural networks, and black box structures in circuits. Topology identification has attracted great attention from various fields, such as inferring connections in multi-agent control systems *Corresponding author (email: hliu@hust.edu.cn) [6], obtaining topologies in distributed energy resources [7], identifying structures in gene co-expression networks [8,9], and inferring relationships in cyber-social networks [10,11]. Therefore, the issue of network structure identification is of great theoretical and practical importance for uncertain complex networks from many fields.…”