Link Prediction with Social Vector Clocks

TL;DR

This paper introduces social vector clocks, a computationally efficient feature for link prediction in social networks, which, when combined with existing methods, achieves state-of-the-art accuracy by leveraging interaction order and timing.

Contribution

The paper proposes social vector clocks as a novel, less expensive feature for link prediction that captures interaction dynamics, improving prediction accuracy.

Findings

Social vector clocks match the performance of complex features.

Combining social vector clocks with existing features yields the most accurate predictions.

Interaction order and spacing are crucial for link formation insights.

Abstract

State-of-the-art link prediction utilizes combinations of complex features derived from network panel data. We here show that computationally less expensive features can achieve the same performance in the common scenario in which the data is available as a sequence of interactions. Our features are based on social vector clocks, an adaptation of the vector-clock concept introduced in distributed computing to social interaction networks. In fact, our experiments suggest that by taking into account the order and spacing of interactions, social vector clocks exploit different aspects of link formation so that their combination with previous approaches yields the most accurate predictor to date.