ConversationMoC

Dataset:

ConversationMoC

Workshop:

AAAI 2024 Workshop W24: Machine Learning for Cognitive and Mental Health Workshop (ML4CHM-2024)

Abstract:

This dataset contains human annotated Reddit post IDs coded for (a) moment of change in user mood and (b) mental health disorder classification. Dataset contains 11,841 unique users, 963 target user conversation timelines over 12 months and a total of 28,659 posts. Post annotations are serialized in JSON format. Post details (text, username, timestamp) will need to be downloaded directly from Reddit via gthe Reddit API using the Reddit post ID's provided. Full details can be found in the paper and readme of the associated github site.

Paper title:

ConversationMoC: Encoding Conversational Dynamics using Multiplex Network for Identifying Moment of Change in Mood and Mental Health Classification

This work introduces a unique conversation-level dataset and investigates the impact of conversational context in detecting Moments of Change (MoC) in individual emotions and classifying Mental Health (MH) topics in discourse. In this study, we differentiate between analyzing individual posts and studying entire conversations, using sequential and graph-based models to encode the complex conversation dynamics. Further, we incorporate emotion and sentiment dynamics with social interactions using a graph multiplex model driven by Graph Convolution Networks (GCN). Comparative evaluations consistently highlight the enhanced performance of the multiplex network, especially when combining reply, emotion, and sentiment network layers. This underscores the importance of understanding the intricate interplay between social interactions, emotional expressions, and sentiment patterns in conversations, especially within online mental health discussions.

This work was supported by the Natural Environment Research Council (NE/S015604/1), the Economic and Social Research Council (ES/V011278/1) and the Engineering and Physical Sciences Research Council (EP/V00784X/1). The authors acknowledge the use of the IRIDIS High Performance Computing Facility, and associated support services at the University of Southampton, in the completion of this work.

Paper DOI:

https://ceur-ws.org/Vol-3649/

Github site:

https://github.com/gyanendrol9/ConversationMOC

Citation:

@inproceedings{
    title = "ConversationMoC: Encoding Conversational Dynamics using Multiplex Network for Identifying Moment of Change in Mood and Mental Health Classification",
    author = "Singh, Loitongbam Gyanendro and Middleton, Stuart E. and Azim, Tayyaba and Nichele, Elena and Lyu, Pinyi and Garcia, Santiago De Ossorno",
    booktitle = "Proceedings of the Machine Learning for Cognitive and Mental Health Workshop (ML4CMH)@AAAI 2024",
    month = Feb,
    year = "2024",
    address = "Vancouver, Canada",
}