4FACTORS Arabic — Egyptian Q&A Sample
Conversational question–answer pairs in spoken Egyptian Arabic, written by a first-language Egyptian speaker. A 50-item demonstration sample, with English glosses, released under CC BY-NC 4.0.
This is the Egyptian variety in the 4FACTORS Arabic sample set, alongside the Palestinian Levantine and Modern Standard Arabic (MSA) sets.
What this is
Fifty short question–answer exchanges of the kind that come up in everyday life — shopping, a doctor's visit, small talk about a football match, arranging a meeting. Each item is a natural spoken-Egyptian question paired with a natural spoken-Egyptian answer, plus a combined English gloss so non-Arabic speakers can evaluate the sample.
- Language / variety: Egyptian Arabic (spoken, colloquial — not MSA)
- Size: 50 Q&A pairs
- Format: CSV, UTF-8
- Glosses: English (
translation_en), combinedquestion / answer - License: CC BY-NC 4.0
Columns
| Column | Description |
|---|---|
id |
Stable item id (eg-001 … eg-050) |
question_ar |
Question in spoken Egyptian Arabic |
answer_ar |
Answer in spoken Egyptian Arabic |
translation_en |
English gloss, question / answer |
domain |
Everyday topic (e.g. Shopping, Health, Transport, Work) |
variety |
Always Egyptian |
How it was made
The data is written by a native speaker — not scraped, not machine-translated, not AI-generated. Every sentence was written by hand by a first-language Egyptian speaker, following a brief that explicitly forbids copying from the web, translating from English, or using any AI tool.
The sample passed two review passes before release:
- Personal data & copyright. Real names, a film title and a restaurant name in the first submission were flagged and replaced with generic equivalents.
- Linguistic diversity. A measured review found the first delivery leaned on a single courtesy formula and long answers; 41 of 100 Q/A cells were reworked by the writer to restore natural variation. One diversity target was deliberately left just short because the native speaker judged the shorter wording unnatural — her judgement was kept over the metric.
The English glosses were drafted by the 4FACTORS team and then reviewed row-by-row by the native Egyptian writer, who confirmed 40 of 50 unchanged and corrected 10.
A full provenance record (chain of custody, before/after review files, checksums) backs this sample.
Intended use
A demonstration of native-written dialectal Arabic data quality — for evaluating fit before commissioning a larger dataset. Suitable for research and non-commercial evaluation. For commercial use or a custom set built to your schema (with full IP transfer and English glosses), contact 4FACTORS: info@4factors.ch.
Licensing & attribution
- License: Creative Commons Attribution-NonCommercial 4.0 (CC BY-NC 4.0).
- Owner: 4FACTORS GmbH (Hugging Face org
4factors, Kaggle orgfourfactors). - The contributor is compensated and has assigned rights in the work product to the company; the sample is published on that basis. No third-party personal data appears in the data.
Archived deposit
This sample is part of an archived corpus (all three varieties, 150 items) published on Zenodo together with the full provenance record — contributor brief, both review passes, the reviewer's correction log and SHA-256 checksums:
Citation
Poljanc, A., & 4FACTORS GmbH (2026). 4FACTORS Arabic Native-Written Sample Corpus:
Modern Standard Arabic, Palestinian Levantine and Egyptian (150 items) (Version 1.0)
[Data set]. Zenodo. https://doi.org/10.5281/zenodo.21380048
- Downloads last month
- 35