Using Proximity to Predict Activity in Social Networks

TL;DR

This paper demonstrates that structural proximity in social networks can effectively predict user activity and URL sharing behavior, improving activity prediction models by incorporating new proximity metrics.

Contribution

The study introduces new proximity metrics for social network analysis and shows their effectiveness in predicting user activity and content sharing.

Findings

Proximity metrics correlate with user activity similarity.

Proximity-based features improve URL recommendation accuracy.

Different proximity metrics vary in predictive performance.

Abstract

The structure of a social network contains information useful for predicting its evolution. Nodes that are "close" in some sense are more likely to become linked in the future than more distant nodes. We show that structural information can also help predict node activity. We use proximity to capture the degree to which two nodes are "close" to each other in the network. In addition to standard proximity metrics used in the link prediction task, such as neighborhood overlap, we introduce new metrics that model different types of interactions that can occur between network nodes. We argue that the "closer" nodes are in a social network, the more similar will be their activity. We study this claim using data about URL recommendation on social media sites Digg and Twitter. We show that structural proximity of two users in the follower graph is related to similarity of their activity, i.e., how many URLs they both recommend. We also show that given friends' activity, knowing their proximity to the user can help better predict which URLs the user will recommend. We compare the performance of different proximity metrics on the activity prediction task and find that some metrics lead to substantial performance improvements.