A Privacy-Preserving Observatory of Misinformation using Linguistic Markers - A Work in Progress

Clos, Jeremie; McClaughlin, Emma; Barnard, Pepita; Tom, Tino; Yajaman, Sudarshan

Abstract

Online misinformation is a serious problem that can have a negative impact on individuals, societies, and democracies. It can lead to the spread of false information, the erosion of trust in institutions, and the polarisation of political discourse. The spread of online misinformation can have a number of negative consequences, including the erosion of trust in institutions, the polarisation of political discourse, and the spread of violence. It can also have a significant impact on individuals, leading to them making decisions based on false information or becoming more isolated from others, leading to wider societal issues. Usual ways of detecting misinformation involve the large scale crawling of public data from public-facing websites (such as social media websites and online news outlets), which is ethically dubious as it violates the right to consent from the online users who generated this data. We present a lightweight, responsible and ethical way of monitoring online misinformation through the use of a distributed approach using a set of linguistic markers extracted from the literature in order to allow users of our browser extension to detect potential misinformation and, with their consent, submit a censored version of that data to our distributed corpus.

Citation

Clos, J., McClaughlin, E., Barnard, P., Tom, T., & Yajaman, S. (2023, July). A Privacy-Preserving Observatory of Misinformation using Linguistic Markers - A Work in Progress. Presented at TAS '23: First International Symposium on Trustworthy Autonomous Systems, Edinburgh, UK

• Presentation Conference Type: Edited Proceedings
• Conference Name: TAS '23: First International Symposium on Trustworthy Autonomous Systems
• Start Date: Jul 11, 2023
• End Date: Jul 12, 2023
• Acceptance Date: Jul 11, 2023
• Online Publication Date: Jul 11, 2023
• Publication Date: Jul 11, 2023
• Deposit Date: Sep 13, 2023
• Publisher: Association for Computing Machinery (ACM)
• Pages: 1-4
• Book Title: TAS '23: Proceedings of the First International Symposium on Trustworthy Autonomous Systems
• ISBN: 9798400707346
• DOI: https://doi.org/10.1145/3597512.3597530
• Public URL: https://nottingham-repository.worktribe.com/output/22726704
• Publisher URL: https://dl.acm.org/doi/10.1145/3597512.3597530