Embedding Mental Health Discourse for Community Recommendation

TL;DR

This paper presents a novel community recommendation system for mental health support groups on social media, leveraging discourse embeddings to improve relevance and interpretability.

Contribution

It introduces an integrated approach combining content-based and collaborative filtering with discourse embeddings for mental health community recommendation.

Findings

Outperforms individual techniques in recommendation accuracy.

Provides interpretability in the recommendation process.

Effective in identifying relevant mental health support groups.

Abstract

Our paper investigates the use of discourse embedding techniques to develop a community recommendation system that focuses on mental health support groups on social media. Social media platforms provide a means for users to anonymously connect with communities that cater to their specific interests. However, with the vast number of online communities available, users may face difficulties in identifying relevant groups to address their mental health concerns. To address this challenge, we explore the integration of discourse information from various subreddit communities using embedding techniques to develop an effective recommendation system. Our approach involves the use of content-based and collaborative filtering techniques to enhance the performance of the recommendation system. Our findings indicate that the proposed approach outperforms the use of each technique separately and provides interpretability in the recommendation process.