{
    "Name": "ArabDiscrim",
    "Dialect Subsets": [
        {
            "Name": "ArabDiscrim",
            "Volume": 293056.0,
            "Unit": "documents",
            "Dialect": "mixed"
        }
    ],
    "HF Link": "",
    "Link": "https://docs.google.com/forms/d/e/1FAIpQLSfnADEgFnSsIz9pDVy4TwpKIGAQRZvbQxU46B8q-rvTpjKaCA/viewform",
    "License": "custom",
    "Year": 2026,
    "Language": "ar",
    "Dialect": "mixed",
    "Source": [
        "social media"
    ],
    "Domain": [
        "society",
        "politics",
        "culture"
    ],
    "Form": "text",
    "Annotation Style": [
        "human validation",
        "metadata-derived"
    ],
    "Description": "A decade-long corpus of 293K Arabic Facebook posts on racism and discrimination.",
    "Volume": 0.0,
    "Unit": "documents",
    "Provider": [
        "Northwestern University in Qatar",
        "Carnegie Mellon University in Qatar"
    ],
    "Derived From": [],
    "Paper Title": "ArabDiscrim: A Decade-Long Arabic Facebook Corpus on Racism and Discrimination",
    "Paper Link": "https://www.lrec-conf.org/proceedings/lrec2026/pdf/2026.lrec2026-1.929.pdf",
    "Script": "Arab",
    "Tokenized": false,
    "Host": "other",
    "Access": "Upon-Request",
    "Cost": "Free",
    "Has Splits": false,
    "Partial": false,
    "Tasks": [
        "text classification",
        "other"
    ],
    "Venue Title": "LREC",
    "Venue Type": "conference",
    "Venue Name": "International Conference on Language Resources and Evaluation",
    "Authors": [
        "Wajdi Zaghouani",
        "Shimaa Amer Ibrahim",
        "Mabrouka Bessghaier",
        "Houda Bouamor"
    ],
    "Affiliations": [
        "Northwestern University in Qatar",
        "Carnegie Mellon University in Qatar"
    ],
    "Abstract": "We present ArabDiscrim, a decade-long lexical resource and corpus of 293K public Arabic Facebook posts (2014\u20132024) discussing racism and discrimination. Unlike existing Twitter-centric datasets, ArabDiscrim integrates platform-native engagement signals, including reactions, shares, comments, and page metadata, enabling joint analysis of language and audience response. The resource includes 200 curated terms (100 racism, 100 discrimination) with morphological regex families (13+ inflections per lemma), and 20 discrimination axes capturing identity-based grounds for unequal treatment. It also provides explicit attribution patterns. Released under a restricted research-use license for ethical compliance with platform terms, ArabDiscrim supports weak supervision, axis-aware sampling, and platform ecology research. By bridging lexical depth and ecological validity, it establishes a foundation for fairness-oriented, platform-aware Arabic NLP.",
    "Added By": "google/gemini-3-flash-preview"
}