Detecting People Interested in Non-Suicidal Self-Injury on Social Media

TL;DR

This paper presents a supervised learning method to identify individuals interested in Non-Suicidal Self-Injury on social media by classifying user interests, validated on LiveJournal data.

Contribution

Introduces a novel binary classification approach using interest-based features to detect NSSI interest on social media platforms.

Findings

Effective classification accuracy demonstrated on LiveJournal dataset

Interest-based features successfully distinguish NSSI interest

Model outperforms baseline methods

Abstract

We propose a supervised learning approach to detect people interested in Non-Suicidal Self-Injury (NSSI). We treat the task as a binary classification problem, and build classifiers based upon features extracted from people self-declared interests. Experimental evaluation on a real-world dataset, the LiveJournal social blogging networking platform, demonstrates the effectiveness of our proposed model.