A survey on proximity measures for social networks

Measuring proximity in a social network is an important task, with many interesting applications, including person search and link prediction. Person search is the problem of finding, by means of keyword search, relevant people in a social network. In user-centric person search, the search query is issued by a person participating in the social network and the goal is to find people that are relevant not only to the keywords, but also to the searcher herself. Link prediction is the task of predicting new friendships (links) that are likely to be added to the network. Both of these tasks require the ability to measure proximity of nodes within a network, and are becoming increasingly important as social networks become more ubiquitous. This chapter surveys recent work on scoring measures for determining proximity between nodes in a social network.We broadly identify various classes of measures and discuss prominent examples within each class. We also survey efficient implementations for computing or estimating the values of the proximity measures.