pith. sign in

arxiv: 2605.23325 · v1 · pith:6F4PFEIYnew · submitted 2026-05-22 · 💻 cs.CL

AraHopeCorpus: Annotation Guidelines and Dataset for Hope Speech in Arabic Social Media Crisis Discourse

Esra'a Sharqawi , Wajdi Zaghouani This is my paper

Pith reviewed 2026-05-25 04:55 UTC · model grok-4.3

classification 💻 cs.CL
keywords hope speechArabicsocial mediacrisis discourseannotated datasetYouTube commentsGaza warinter-annotator agreement
0
0 comments X

The pith

AraHopeCorpus is the first annotated Arabic dataset of hope speech from ten thousand YouTube comments on the Gaza war, where hopeful expressions exceed sixty four percent.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper creates and releases AraHopeCorpus, an annotated collection of Arabic comments from social media during the 2023-2024 Gaza conflict. Comments are labeled as hope speech, no hope speech, or neutral, with hope speech making up more than sixty four percent of the total. The dataset highlights common forms of hope like religious encouragement and collective solidarity. It also compares human annotation to large language models and finds the models struggle with dialectal Arabic. The resource supports further study of positive communication in crisis settings.

Core claim

The paper presents AraHopeCorpus as the first such resource for Arabic, collected from crisis-related YouTube comments, and demonstrates through annotation that hopeful language is the most common category at over sixty four percent, primarily through religious, solidarity-based, and optimistic expressions.

What carries the argument

AraHopeCorpus, a dataset of ten thousand annotated Arabic YouTube comments categorized into hope speech, no hope speech, and neutral or unclear, along with the accompanying annotation guidelines.

Load-bearing premise

The three annotation categories can be applied consistently to informal, dialectal Arabic text even when sarcasm or implicit cultural references are present.

What would settle it

Re-annotating a random sample of the comments with new annotators and obtaining substantially lower agreement scores or a different category distribution.

Figures

Figures reproduced from arXiv: 2605.23325 by Esra'a Sharqawi, Wajdi Zaghouani.

Figure 2
Figure 2. Figure 2: Word cloud of hope speech frequent words(Voyant tools) The hope speech category, which formed the majority of the dataset, was dominated by religious and spiritual encouragement (50.48%), reflected in frequent references to -.1JJ (Allah), �•1JJ (O Allah), and __ l, (O Lord), illustrating faith as a key source of resilience. This was followed by support and encouragement (33.68%), where mentions of [PITH_F… view at source ↗
read the original abstract

Social media has become a crucial arena for shaping public narratives during armed conflicts, providing space for both harmful and constructive communication. While hate speech and misinformation have been widely studied, expressions that promote resilience, solidarity, and optimism remain underexplored, particularly in Arabic contexts. This paper introduces AraHopeCorpus, the first annotated dataset of Arabic hope speech collected from ten thousand YouTube comments related to the war on Gaza between 2023 and 2024. Using a detailed annotation framework, comments were classified into three categories: hope speech, no hope speech, and neutral or unclear discourse. The dataset shows that hopeful language dominates, accounting for more than sixty four percent of all comments. These expressions of hope appear mainly as religious encouragement, collective solidarity, and optimism for endurance and justice. No hope speech, representing about thirteen percent, reflects despair and disillusionment, while the rest of the comments contain neutral or mixed content. Inter-Annotator Agreement reached substantial levels (Cohen's Kappa equals 0.71), though dialectal variation, sarcasm, and implicit meaning posed annotation challenges. A comparative analysis between human annotators and ChatGPT revealed that large language models can support annotation but remain limited in handling dialectal and culturally embedded expressions. AraHopeCorpus will be released for research purposes under an open and non commercial license. It provides a valuable resource for studying constructive digital discourse, enabling further research on hope speech detection, crisis communication, and resilience in Arabic social media.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper introduces AraHopeCorpus, the first annotated dataset of Arabic hope speech from 10,000 YouTube comments on the 2023-2024 Gaza war. Comments are labeled into three categories (hope speech >64%, no hope speech ~13%, neutral/unclear) using explicit guidelines; Cohen's Kappa reaches 0.71. The work includes annotation challenges, a human-LLM comparison, and plans to release the corpus under an open non-commercial license.

Significance. If the annotations hold, the resource fills a clear gap in non-English hope-speech research during crises and supports downstream work on constructive discourse detection. The reported IAA, category definitions, and release commitment are direct strengths that enable community use.

minor comments (3)
  1. [Data Collection] Data collection section: specify the exact keywords, channels, and sampling procedure used to obtain the 10,000 comments so that the corpus can be replicated or extended.
  2. [Comparative Analysis] LLM comparison section: state the exact prompt template and temperature settings given to ChatGPT; without them the comparative results cannot be reproduced.
  3. [Results] Table 1 (category distribution): add row percentages with 95% confidence intervals to quantify uncertainty around the reported 64% and 13% figures.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive evaluation of the manuscript and the recommendation to accept. The review accurately captures the contribution of AraHopeCorpus as the first annotated Arabic dataset for hope speech in crisis-related social media discourse.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper is an empirical contribution that introduces AraHopeCorpus by collecting and annotating 10,000 YouTube comments into three categories, reporting observed distributions (>64% hope speech) and Cohen's Kappa of 0.71. No equations, parameter fitting, derivations, predictions, or self-citation chains exist that could reduce any claim to its own inputs by construction. The annotation process is validated against external human judgments rather than internally forced, and the work is transparent about dialectal and cultural challenges without claiming generalizability beyond the sample.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper rests on the domain assumption that hope speech can be distinguished from despair and neutral content in Arabic social media text via human annotation; no free parameters or invented entities are introduced.

axioms (1)
  • domain assumption Standard practices for defining and applying hope/no-hope/neutral categories to social media comments produce reliable labels when inter-annotator agreement reaches substantial levels.
    Invoked in the description of the annotation framework and Cohen's Kappa result.

pith-pipeline@v0.9.0 · 5800 in / 1308 out tokens · 21244 ms · 2026-05-25T04:55:58.653806+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages

  1. [1]

    El-Hardlo, and F

    Ines Abbes, Wajdi Zaghouani, O. El-Hardlo, and F. Achour. 2020. Daict: A dialectal arabic irony corpus extracted from twitter. In Proceedings of the Language Resources and Evaluation Confer- ence. A. Abdelali et al. 2020. Protest discourse in arabic social media. In Proceedings of WANLP. M. Al Emadi and Wajdi Zaghouani. 2024. Emotional toll and coping str...

    work page 2020