摘要：Abstract Two computational methods based on the Ising model were implemented for studying temporal dynamic in co-authorship networks: an interpretative for real networks and another for simulation via Monte Carlo. The objective of simulation networks is to evaluate if the Ising model describes in similar way the dynamic of the network and of the magnetic system, so that it can be found a generalized explanation to the behaviours observed in real networks. The scientific papers used for building the real networks were acquired from WoS core collection. The variables for each record took into account bibliographic references. The search equation for each network considered specific topics trying to obtain an advanced temporal evolution in terms of the addition of new nodes; that means 3 steps, a time to reach the interest of the scientific community, a gradual increase until reaching a peak and finally, a decreasing trend by losing of novelty. It is possible to conclude that both methods are consistent with each other, showing that the Ising model can predict behaviours such as the number and size of communities (or domains) according to the temporal distribution of new nodes.
其他摘要：Abstract Two computational methods based on the Ising model were implemented for studying temporal dynamic in co-authorship networks: an interpretative for real networks and another for simulation via Monte Carlo. The objective of simulation networks is to evaluate if the Ising model describes in similar way the dynamic of the network and of the magnetic system, so that it can be found a generalized explanation to the behaviours observed in real networks. The scientific papers used for building the real networks were acquired from WoS core collection. The variables for each record took into account bibliographic references. The search equation for each network considered specific topics trying to obtain an advanced temporal evolution in terms of the addition of new nodes; that means 3 steps, a time to reach the interest of the scientific community, a gradual increase until reaching a peak and finally, a decreasing trend by losing of novelty. It is possible to conclude that both methods are consistent with each other, showing that the Ising model can predict behaviours such as the number and size of communities (or domains) according to the temporal distribution of new nodes.
