Modified Heider Balance on Sparse Random Networks
The lack of signed random networks in standard balance studies has prompted us to extend the Hamiltonian of the standard balance model. Random networks with tunable parameters are suitable for better understanding the behavior of standard balance as an underlying dynamics. Moreover, the standard balance model in its original form does not allow preserving tensed triads in the network. Therefore, the thermal behavior of the balance model has been investigated on a fully connected signed network recently. It has been shown that the model undergoes an abrupt phase transition with temperature. Considering these two issues together, we examine the thermal behavior of the structural balance model defined on Erd\H{o}s-R\'enyi random networks. We provide a Mean-Field solution for the model. We observe a first-order phase transition with temperature, for both the sparse and densely connected networks. We detect two transition temperatures, $T_{cold}$ and $T_{hot}$, characterizing a hysteresis loop. We find that with increasing the network sparsity, both $T_{cold}$ and $T_{hot}$ decrease. But the slope of decreasing $T_{hot}$ with sparsity is larger than the slope of decreasing $T_{cold}$. Hence, the hysteresis region gets narrower, until, in a certain sparsity, it disappears. We provide a phase diagram in the temperature-tie density plane to observe the meta-stable/coexistence region behavior more accurately. Then we justify our Mean-Field results with a series of Monte-Carlo simulations.