Game-Theoretic Approaches to Community Detection in Social Networks

Game-theoretic methods frame the problem of community detection as a strategic interaction among nodes, each regarded as a rational agent seeking to optimise an individual or collective payoff. Approaches span non-cooperative frameworks, in which nodes independently choose community affiliations to maximise personal utility, and cooperative models, where nodes form coalitions and share rewards according to a characteristic function. Evolutionary dynamics and optimisation techniques guide nodes towards equilibrium partitions that reveal both disjoint and overlapping communities. Metrics such as modularity and stability of Nash equilibria underpin algorithmic design, while advances in convex game theory and submodular function optimisation guarantee unique and interpretable partitions. These methods address scalability through linear or polynomial-time algorithms, accommodate network evolution, and tolerate noise and incomplete data. Their applications range from tracking shifting opinion clusters and countering misinformation in social media to enhancing recommendation systems and understanding organisational behaviour. Interdisciplinary insights from economics, computer science and statistical physics continue to enrich this field, offering robust theoretical foundations and practical tools for large-scale social networks.

Research from Nature Portfolio

A recent non-cooperative framework casts each node as a strategic player selecting a community label to maximise individual utility. The two-phase design first identifies candidate communities and then refines overlapping memberships, eliminating the need for preset parameters on community size or number. The absence of stochastic elements ensures deterministic convergence and linear-time complexity, making the algorithm scalable to large networks. Experiments on synthetic and real-world datasets demonstrate high modularity and stability in detected structures, with particular strength in unveiling overlapping community cores. This work advances both theoretical understanding of Nash equilibria in community assignment and practical deployment in extensive social platforms.

Game-Theoretic Approaches to Community Detection in Social Networks publication trend

The graph below shows the total number of articles in game-theoretic approaches to community detection in social networks across all publications each year (not limited to Nature Index journals).

Technical terms

Community detection: process of identifying groups of nodes with dense internal and sparse external connections.

Non-cooperative game: framework in which each player independently selects a strategy to maximise individual payoff.

Cooperative game: approach where players form coalitions and share collective payoffs based on a characteristic function.

Evolutionary game: dynamic process in which strategies adapt over time according to relative payoffs.

Shapley value: solution concept that allocates total coalition gains fairly among players based on marginal contributions.

Modularity: quality metric comparing intra-community link density with a random null model.

Submodular function minimisation: optimisation technique for functions exhibiting diminishing returns, used to derive unique partitions.

Nash equilibrium: stable state in a non-cooperative game where no player can improve payoff by unilateral deviation.

References

1. Game Theory and Extremal Optimization for Community Detection in Complex Dynamic Networks. PLOS ONE (2014).
2. A Cooperative Game Theory-Based Algorithm for Overlapping Community Detection. IEEE Access (2020).
3. Detecting overlapping communities in complex networks using non-cooperative games. Scientific Reports (2022).
4. Game Theoretic Clustering for Finding Strong Communities. Entropy (2024).