Research on social networks is a fashionable field, along with forecasting the behaviors of entities that are represented as nodes in a graph describing the relationship between entities. The authors investigate opportunities for the improvement of algorithms that were developed previously and tackled the problem of how to predict the creation of new edges between already-existing nodes, taking into account the time and adequate probability distribution.

The paper contains a comprehensive literature survey that includes the direct antecedent of the proposed approach: the co-occurrence probability measure. The authors extend the original definition into a so-called temporal co-occurrence probability measurement. An algorithm is described for link prediction in networks, using the previously defined probabilistic measure.

The subject of the empirical investigation of the goodness of the proposed algorithm is a dataset that consists of databases containing publications and citations in various scientific disciplines. The underlying idea of the proposed approach and algorithm is that the clique of graphs--that is, complete sub-graphs--should be sifted and then probability measures calculated for each clique to construct a Markov random field for the cliques. The basic idea is that two arbitrary nodes are selected, all of the cliques that contain the selected pair of nodes are collected, and then a joint probability for the selected nodes is calculated.

The evaluation measures for the proposed algorithm are the widely accepted area under receiver operating characteristic (AUROC) and area under the precision-recall curve (AUPR). The assessment is carried out on two dense and two sparse scientific databases. The presented results show that the proposed algorithm works better than other state-of-the-art algorithms because it considers the age of cliques as an important temporal factor. The authors claim that the developed algorithm will operate well in social networks. This is an interesting paper for researchers in the fields of social networks and data science.