Link Prediction by De-anonymization: How We Won the Kaggle Social Network Challenge

TL;DR

This paper details how de-anonymization techniques were used to win a social network link prediction challenge, highlighting vulnerabilities in privacy and proposing a combined approach for improved prediction accuracy.

Contribution

The paper introduces a novel de-anonymization attack for social network graphs and integrates it with link prediction methods to enhance performance in a competitive setting.

Findings

Successfully de-anonymized a significant portion of the test set

Combined de-anonymization with link prediction for better accuracy

Highlighted privacy vulnerabilities in social network data

Abstract

This paper describes the winning entry to the IJCNN 2011 Social Network Challenge run by Kaggle.com. The goal of the contest was to promote research on real-world link prediction, and the dataset was a graph obtained by crawling the popular Flickr social photo sharing website, with user identities scrubbed. By de-anonymizing much of the competition test set using our own Flickr crawl, we were able to effectively game the competition. Our attack represents a new application of de-anonymization to gaming machine learning contests, suggesting changes in how future competitions should be run. We introduce a new simulated annealing-based weighted graph matching algorithm for the seeding step of de-anonymization. We also show how to combine de-anonymization with link prediction---the latter is required to achieve good performance on the portion of the test set not de-anonymized---for example by training the predictor on the de-anonymized portion of the test set, and combining probabilistic predictions from de-anonymization and link prediction.