There are many metrics that can be used in a link prediction problem. We have already studied some of these in this book but we will review them with a new focus on link prediction in this section. Some other metrics have been introduced especially for this kind of application and come under the linkprediction namespace in the GDS.
The idea of link prediction algorithms is to be able to create a matrix N×N, where N is the number of nodes in the graph. Each ij element of the matrix must give an indication of the probability of the existence of a link between nodes i and j.
Different kinds of metrics can be used to achieve this goal. One of these is node similarity metrics, such as the Jaccard similarity we studied in Chapter 7, Community Detection and Similarity Measures. In this method, by comparing the set of node neighbors, we can get an idea about the nodes' similarities and how likely they are to be connected in the future.
Similarity...
