Who Should Set the Standards? Analysing Censored Arabic Content on Facebook during the Palestine-Israel Conflict

TL;DR

The paper investigates fairness in Facebook’s moderation of Arabic content during the Palestine-Israel conflict, revealing a substantial gap between Arab users’ perceptions and FB’s implementation of its Community Standards. Using 448 deleted Arabic posts assessed by 10 Arab annotators per item, the study finds that a majority of posts do not violate the FBCS according to local users, and most should not be removed based on personal opinion, while hate-speech content is more consistently flagged. The results raise questions about who should set and interpret global moderation guidelines, emphasizing cultural sensitivity, transparency, and contestability to improve legitimacy and inclusivity. The work highlights practical implications for HCI researchers and social platforms to incorporate diverse voices and context-aware governance in content moderation to better protect digital rights across communities.

Abstract

Nascent research on human-computer interaction concerns itself with fairness of content moderation systems. Designing globally applicable content moderation systems requires considering historical, cultural, and socio-technical factors. Inspired by this line of work, we investigate Arab users' perception of Facebook's moderation practices. We collect a set of 448 deleted Arabic posts, and we ask Arab annotators to evaluate these posts based on (a) Facebook Community Standards (FBCS) and (b) their personal opinion. Each post was judged by 10 annotators to account for subjectivity. Our analysis shows a clear gap between the Arabs' understanding of the FBCS and how Facebook implements these standards. The study highlights a need for discussion on the moderation guidelines on social media platforms about who decides the moderation guidelines, how these guidelines are interpreted, and how well they represent the views of marginalised user communities.

Figures (6)

• Figure 1: Rating of the Facebook app in Google's Play Store (screenshots). Previously, the rating was 4.0 (out of 5), but the activist campaign decreased the rating in May 2021. Screenshot (a) shows the rating in 17 May 2021, few days after the start of the campaign. Screenshot (b) and (c) is from September 9, 2021 and August 16, 2023 respectively, implying that the damage to reputation from activist campaigns can be long-lasting.
• Figure 2: Demographics and account reach of the authors of the 448 deleted Facebook posts
• Figure 3: Statistics on the nature of the posts (a), restriction applied to the account (b), and appeal on restriction for the 448 posts (c).
• Figure 4: Distribution of the topics discussed in the collected 448 posts.
• Figure 5: Percentage of posts that should be removed based on the minimum number of annotators who agree that they violate the FBCS or according to their personal opinion.