Social governance is a crucial component of national governance and an important foundation for maintaining social stability, promoting public welfare, and improving governance capacity. In recent years, the world has faced increasingly complex governance challenges, including poverty, public health crises, natural disasters, and large-scale population mobility. Traditional governance approaches, which largely rely on census statistics, field surveys, and empirical judgment, often suffer from limited timeliness, insufficient spatial granularity, and inadequate dynamic monitoring capability. With the rapid development of mobile communication networks, satellite positioning systems, Internet platforms, and Internet of Things technologies, massive volumes of spatiotemporal behavioral data are continuously generated through human mobility activities. Compared with traditional statistical data, mobile big data possesses advantages in coverage, continuity, temporal resolution, and acquisition cost, making it possible to characterize population movement, social interaction, and spatial activity patterns in near real time. Consequently, mobile big data has become an important technological foundation for promoting the transformation of social governance from experience-driven decision-making to data-driven intelligent governance. Systematically reviewing the research progress and application scenarios of mobile big data is therefore of great significance for advancing precision governance, emergency management, public service optimization, and digital governance modernization.

This paper systematically reviewed the major types, methodological progress, and representative applications of mobile big data in social governance. First, the paper summarized the fundamental characteristics and data sources of mobile big data. In addition to the widely recognized “5V” characteristics of big data, the paper further proposed the “5C” characteristics of mobile big data, namely comprehensiveness, continuity, chrono-spatiality, confidentiality sensitivity, and citizen-centricity. According to differences in positioning technologies, data ownership, and application scenarios, mobile big data was categorized into four major types: mobile phone positioning data, public transportation data, Internet user-authorized data, and Internet of Things pervasive sensing data. Their respective advantages, limitations, and potential biases in social governance applications were comparatively analyzed.

The paper further reviewed research progress in human mobility behavior mining and mobility mechanism modeling. Existing studies demonstrated that human mobility exhibited universal statistical regularities, including power-law displacement distributions, periodic return behaviors, high predictability, and strong spatiotemporal regularity. Representative studies based on mobile phone trajectory data revealed that human movement patterns were neither completely random nor fully deterministic, but instead followed stable behavioral mechanisms constrained by spatial hierarchy, social relationships, and memory effects. On this basis, a series of mobility mechanism models were proposed, including gravity models, intervening opportunity models, radiation models, extended radiation models, and preferential exploration and preferential return models. These models gradually evolved from empirical parameterized approaches toward theoretically interpretable and scalable frameworks capable of bridging individual mobility behavior and macroscopic population flow patterns. Their advantages and limitations under different governance scenarios were systematically compared in terms of parameter dependence, spatial adaptability, and robustness to structural changes.

The paper also reviewed advances in mobility network modeling. Mobility processes were abstracted into travel networks and contact networks by representing locations or individuals as nodes and mobility interactions as edges. Research on travel networks explored urban structure identification, commuting analysis, transportation resilience, and multilayer network coupling, while contact network studies focused on epidemic transmission, social interaction structures, and high-order network dynamics. Recent studies further incorporated temporal evolution and high-order interactions into mobility network models, enabling more accurate characterization of cascading failures, collective behavioral changes, and dynamic spreading processes under complex governance environments.

Finally, the paper systematically reviewed representative applications of mobile big data in social governance. In poverty identification, researchers utilized communication records, mobility indicators, and social interaction features to infer regional economic conditions and generate high-resolution poverty maps, particularly in data-scarce developing regions. In economic assessment, mobility intensity, nighttime activity, transportation flows, and commercial visitation patterns were employed to estimate economic vitality, urban prosperity, and regional development levels. In epidemic prevention and control, mobile big data played a key role in monitoring population movement, modeling disease transmission, evaluating intervention effectiveness, and supporting dynamic policy adjustment during the COVID-19 pandemic. In emergency response and disaster management, mobile big data was used to monitor evacuation behavior, estimate affected populations, optimize rescue resource allocation, and evaluate infrastructure resilience during earthquakes, floods, typhoons, and other emergencies. These studies demonstrate the strong practical value of mobile big data in supporting real-time situational awareness and intelligent governance decision-making.

Mobile big data has significantly expanded the capability of social governance by providing fine-grained, continuous, and large-scale observations of human mobility and social interaction. Existing studies have established relatively mature theoretical foundations in mobility behavior analysis, network modeling, and governance-oriented applications, demonstrating the broad applicability of mobile big data in poverty reduction, economic monitoring, epidemic control, and emergency management. Nevertheless, several critical challenges remain unresolved. First, data representativeness and sampling bias continue to affect the reliability and fairness of governance decisions, especially for populations with limited digital access. Second, mobility data inherently contains sensitive personal information, and issues concerning privacy protection, data security, and legal compliance require further attention. Third, differences in data standards, spatial scales, and semantic representations hinder effective integration across heterogeneous data sources.

Future research should focus on several directions. One important direction is the integration and semantic alignment of multi-source spatiotemporal data to improve the completeness and robustness of governance analytics. Another direction involves the development of privacy-preserving computation frameworks and federated learning techniques that enable secure data utilization without exposing individual trajectories. In addition, real-time disaster response systems based on streaming mobile data and adaptive network modeling will become increasingly important for improving urban resilience. Furthermore, with the rapid progress of artificial intelligence and large language models, future studies are expected to combine mobile big data with artificial society simulation and digital twin technologies, thereby enabling large-scale social behavior prediction, policy evaluation, and intelligent governance simulation. These developments will further promote the evolution of social governance toward more precise, adaptive, and intelligent paradigms.