Decentralised Moderation for Interoperable Social Networks: A Conversation-Based Approach for Pleroma and the Fediverse
Abstract
The recent development of decentralised and interoperable social networks (such as the "fediverse") creates new challenges for content moderators. This is because millions of posts generated on one server can easily "spread" to another, even if the recipient server has very different moderation policies. An obvious solution would be to leverage moderation tools to automatically tag (and filter) posts that contravene moderation policies, e.g. related to toxic speech. Recent work has exploited the conversational context of a post to improve this automatic tagging, e.g. using the replies to a post to help classify if it contains toxic speech. This has shown particular potential in environments with large training sets that contain complete conversations. This, however, creates challenges in a decentralised context, as a single conversation may be fragmented across multiple servers. Thus, each server only has a partial view of an entire conversation because conversations are often federated across servers in a non-synchronized fashion. To address this, we propose a decentralised conversation-aware content moderation approach suitable for the fediverse. Our approach employs a graph deep learning model (GraphNLI) trained locally on each server. The model exploits local data to train a model that combines post and conversational information captured through random walks to detect toxicity. We evaluate our approach with data from Pleroma, a major decentralised and interoperable micro-blogging network containing 2 million conversations. Our model effectively detects toxicity on larger instances, exclusively trained using their local post information (0.8837 macro-F1). Yet, we show that this approach does not perform well on smaller instances that do not possess sufficient local training data. Thus, in cases where a server contains insufficient data, we strategically retrieve information (posts or model parameters) from other servers to reconstruct larger conversations and improve results. With this, we show that we can attain a macro-F1 of 0.8826. Our approach has considerable scope to improve moderation in decentralised and interoperable social networks such as Pleroma or Mastodon.Dataset and Codes
An anonymized version of the dataset and codes used in our paper are available for the research community.
Pleroma Dataset: This dataset contains decentralised conversation trees for Pleroma.
-
Codes: Implementation details and codes for GraphNLI are available on GITHUB.
You can find the format of the Pleroma dataset here.
Contact Us
If you are interested in using this data, please fill the form to . Request specific data to get the link where you can download the data.
We are sharing the dataset under the terms and conditions specified here below. Please note that submitting the form indicates that you accept the terms and conditions of the data. In the form, please indicate which part of the dataset you need. If you do not get any email notification for your logged request within 24 hours, please e-mail us at netsys.noreply[at]gmail.com.Dataset Terms and Conditions
You will use the data solely for the purpose of non-profit research or non-profit education.
You will respect the privacy of end users and organizations that may be identified in the data. You will not attempt to reverse engineer, decrypt, de-anonymize, derive or otherwise re-identify anonymized information.
You will not distribute the data beyond your immediate research group.
If you create a publication using our datasets, please cite our papers as follows.
@inproceedings{agarwal2024decentralised,
title={Decentralised Moderation for Interoperable Social Networks: A Conversation-based Approach for Pleroma and the Fediverse},
author={Agarwal, Vibhor and Raman, Aravindh and Sastry, Nishanth and Abdelmoniem, Ahmed M and Tyson, Gareth and Castro, Ignacio},
booktitle={Proceedings of the International AAAI Conference on Web and Social Media},
volume={18},
pages={2--14},
year={2024}
}