Abstract:Consider spies infiltrating a network or dissidents secretly organising under a dictatorship. Such scenarios can be cast as adversarial social network analysis problems involving nodes connecting while evading network analysis tools, e.g., centrality measures or community detection algorithms. While most works consider unilateral actions of an evader, we define a network formation game. Here, several newcomers attempt to rewire the existing social network, to become directly tied with the high centrality players, while keeping their own centrality small. This extends the network formation literature, including the Jackson and Wolinsky model, by considering additional strategies and new utility functions. We algorithmically demonstrate that the pairwise Nash stable networks (\PANS) constitute a lattice, where the stronger \PANS{} lattice is nested in the weaker \PANS. We also prove that inclusion in \PANS{} implies less utility for everyone. Furthermore, we bound the social efficiency of \PANS{} and directly connect efficiency to the strength of \PANS. Finally, we characterise the \PANS{} in practically important settings, deriving tight efficiency bounds. Our results suggest the hiders how to interconnect stably and efficiently. Additionally, the results let us detect infiltrated networks, enhancing the social network analysis tools. Besides the theoretical development, this is applicable to fighting terrorism and espionage.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how multiple evaders (such as spies or dissidents) can maintain a low centrality of their own while connecting to high - centrality nodes in adversarial social network analysis. Specifically, the paper studies how multiple new entrants can reconnect an existing social network in the framework of network formation games to directly establish connections with high - centrality players while keeping their own centrality small. ### Problem Background In many cases, agents in the network wish to hide their cooperation and participation. For example: - Terrorists attempt to conceal their leadership structure from security forces. - Political dissidents hide themselves under a dictatorship. - Lobbying for foreign or unpopular interests requires secrecy. - Unions and economic cooperators may initially wish to avoid publicity. - Software agents may choose to hide for reasons of confidentiality or to avoid direct connection with potentially harmful agents. These scenarios can be modeled as adversarial social network analysis problems, in which network members evade detection by social network analysis tools (such as centrality measures or community detection algorithms) by reconnecting the network. ### Main Contributions of the Paper 1. **Introducing a New Network Formation Game Model**: For the first time, the paper models the adversarial social network analysis problem as a network formation game among multiple evaders. In this game, new entrants attempt to connect to or reconnect an existing social network to directly establish connections with high - centrality nodes while keeping their own centrality small. 2. **Discovering the Lattice Structure of the Stable State**: The paper proves that pairwise Nash - stable networks (PANS) form a lattice, and stronger PANS lattices are nested within weaker PANS. This provides important insights into the network topology in the stable state. 3. **Efficiency Analysis**: The paper analyzes the efficiency of PANS and relates it to the strength of PANS. The author also provides tight efficiency bounds for specific scenarios. 4. **Practical Applications**: The research results not only contribute to theoretical development but can also be applied to combating terrorism and espionage and enhancing social network analysis tools. ### Summary of Mathematical Formulas - **Utility Function**: \[ u_i(s)=\sum_{j \in N(i)} \text{deg}(j)-\alpha_i\cdot \text{deg}(i) \] where \(N(i)\) is the set of neighbors of player \(i\), \(\text{deg}(i)\) is the degree of player \(i\), and \(\alpha_i\) is the cost of publicity for player \(i\). - **Pairwise Nash - stable Networks (PANS)**: - If the utilities of both players \(i\) and \(j\) increase after adding the edge \((i, j)\), then these two players should be connected. - PANS form a lattice, and stronger PANS are nested within weaker PANS. ### Conclusion By introducing a new network formation game model, the paper extends the existing adversarial social network analysis and discovers the lattice structure in the stable state. This not only provides a new perspective for theoretical research but also offers valuable insights for practical applications (such as combating terrorism and espionage).

Hiders' Game

Social Distancing Network Creation

Efficient Best-Response Computation for Strategic Network Formation under Attack

Network Creation with Homophilic Agents

The Hitchhiker's Guide to Affiliation Networks: A Game-Theoretic Approach

Adversarial Classification on Social Networks

Deception Maze: A Stackelberg Game-Theoretic Defense Mechanism for Intranet Threats

Honeypot Allocation for Cyber Deception in Dynamic Tactical Networks: A Game Theoretic Approach

Adaptive Payoff-driven Interaction in Networked Snowdrift Games

On the complexity of sabotage games for network security

The topology of covert conflict

Network, Popularity and Social Cohesion: A Game-Theoretic Approach

Steering cooperation: Adversarial attacks on prisoner's dilemma in complex networks

Network partitioning algorithms as cooperative games

Implication Avoiding Dynamics for Externally Observed Networks

Network Formation under Random Attack and Probabilistic Spread

Network Members Can Hide from Group Centrality Measures

Sharp thresholds limit the benefit of defector avoidance in cooperation on networks

Adversarial Coordination on Social Networks

A two-network adversarial game: Model, strategy, and structure

Adversaries with Limited Information in the Friedkin--Johnsen Model