Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Table 1.

Payoff matrix of the Prisoner's Dilemma.

More »

Table 1 Expand

Table 2.

Statistical properties of e-mail and PGP networks.

More »

Table 2 Expand

Figure 1.

Evolution of cooperation in real social networks.

Black lines: Density of cooperators as a function of b, obtained by numerical simulations on the email (left) and PGP (right) networks. Red lines: Density of cooperators on random networks generated from the original ones by a rewiring procedure that preserves the degree distribution(see text). The equilibrium densities of cooperators have been obtained by averaging 500 generations, after a transient time of 750 generation steps. Each point corresponds to an average over 1000 independent simulations with 50% cooperators and defectors as the initial condition.

More »

Figure 1 Expand

Figure 2.

Community structures of the email and PGP networks.

Top: Community structures of the email (A) and PGP (B) networks. Nodes correspond to communities (where size is proportional to their number of members) and links represent cross-connections (where width corresponds to the number of inter-connetions). Bottom: Typical examples of the communities detected in the email (C) and PGP (D) networks. Solid links join nodes of the community, dashed links join this community with others.

More »

Figure 2 Expand

Table 3.

Statistical properties of synthetic networks with 10000 nodes and 75 communities.

More »

Table 3 Expand

Figure 3.

Evolution of cooperation in four synthetic networks.

Cases A and D correspond, respectively, to the synthetic classes of networks akin to the email and PGP real networks. In case A communities have been built as Erdos-Renyi random graphs (pintra = 1.5×10−1), and the probability of interconnection between communities (pinter) is 5×10−2. Communities in case D are constructed as independent scale-free networks (Barabasi-Albert with k0 = 3), and after they have been sparsely interconnected with (pinter = 1.5×10−5). Case B has been obtained from D by increasing the probability pinter to 3.5×10−4, and case C corresponds to A reducing this probability to 7.5×10−4. Simulations have been performed as indicated in Fig. 1.

More »

Figure 3 Expand