<ul data-eligibleForWebStory="true"><li>Malicious sockpuppet detection on Wikipedia is crucial to maintaining reliable information and combating disinformation.</li><li>Previous machine learning methods focused on style and metadata features but lacked adaptability to specific author behaviors.</li><li>These approaches struggled to model the behavior of certain sockpuppet groups with limited text data.</li><li>To improve sockpuppet detection, meta-learning, a technique enhancing performance in data-scarce scenarios, is proposed.</li><li>Meta-learning allows models to quickly adapt to new sockpuppet-group writing styles.</li><li>Results indicate that meta-learning boosts prediction accuracy compared to pre-trained models, a step forward in countering sockpuppetry on open platforms.</li><li>A new dataset of sockpuppet investigations has been released to support further research in sockpuppetry and meta-learning domains.</li></ul>

Detecting Sockpuppetry on Wikipedia Using Meta-Learning

Discover more