Graph-Based Algorithms for Social Network Analysis

Graph-based algorithms form the backbone of contemporary social network analysis by modelling individuals or entities as vertices and their relationships as edges in a graph. This abstraction enables rigorous study of network topology, information flow, influence dynamics and community structure. Key tasks include centrality measurement, which identifies influential nodes using walk-based or spectral methods; link prediction, which anticipates future relationships via similarity metrics or probabilistic models; and community detection, which uncovers tightly connected subgroups. Community detection often relies on clique-based definitions and their relaxations—such as k-plexes and s-clubs—to capture realistic, overlapping group structures. Advances in combinatorial optimisation, parameterised complexity and exact algorithms now allow the enumeration of significant subgraphs in networks of hundreds of thousands of nodes. At the same time, heuristic and approximation techniques enable scalable analysis of platforms with millions of users. Recent work has also integrated graph-theoretic methods with machine-learning approaches, notably graph neural networks, to combine structural insights with attribute data. These developments underpin applications in public health for disease propagation modelling, in marketing for influence maximisation, and in security for fraud detection, demonstrating the global significance and practical utility of graph-based analysis.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent algorithmic innovations have significantly improved the detection of cohesive substructures in large-scale networks. A new branch-and-bound method for triangle-constrained 2-clubs handles diameter-two communities where each node participates in a minimum number of triangles, achieving exact solutions on real-world graphs with over 100 000 vertices in minutes. In parallel, studies on the parameterised complexity of s-Club variants have established fixed-parameter tractable algorithms for cases with triangle or seed constraints, clarifying which community definitions admit efficient exact solutions despite underlying NP-hardness. Complementing these exact techniques, a meta-algorithm for large k-plex enumeration selectively filters out insignificant subgraphs, enabling orders-of-magnitude speed-up in the enumeration of near-clique structures. Together, these works demonstrate a trend towards hybridising theoretical complexity analyses with practical reductions and heuristics, yielding tools that can be tuned to the size, density and specific cohesion criteria of diverse social networks.

Graph-Based Algorithms for Social Network Analysis publication trend

The graph below shows the total number of articles in graph-based algorithms for social network analysis across all publications each year (not limited to Nature Index journals).

Technical terms

Clique: A subgraph in which every pair of vertices is connected by an edge, representing maximally cohesive groups.

k-plex: A relaxation of a clique allowing each vertex to be non-adjacent to at most k other vertices, capturing near-clique communities.

2-club: A subgraph whose diameter does not exceed two, modelling communities in which any two members are connected by at most one intermediary.

s-Club: A generalisation of a 2-club in which the maximum distance between any two vertices is at most s.

Branch-and-bound algorithm: An exact optimisation method that systematically explores candidate solutions and prunes suboptimal branches based on bounding functions.

Fixed-parameter tractability: A complexity paradigm where computationally hard problems can be solved efficiently for small values of chosen parameters despite their NP-hard status.

References

1. The Parameterized Complexity of s-Club with Triangle and Seed Constraints. Theory of Computing Systems (2023).
2. A meta-algorithm for finding large k-plexes. Knowledge and Information Systems (2021).
3. Efficient branch-and-bound algorithms for finding triangle-constrained 2-clubs. Journal of Combinatorial Optimization (2024).