The MADAR Arabic Dialect Corpus and Lexicon

By Houda Bouamor, Nizar Habash, Mohammad Salameh, W. Zaghouani, Owen Rambow, Dana Abdulrahim, Ossama Obeid, Salam Khalifa, Fadhl Eryani, Alexander Erdmann and 1 more. International Conference on Language Resources and Evaluation, 2018

296 citations recorded by the source corpus.

In this paper, we present two resources that were created as part of the Multi Arabic Dialect Applications and Resources (MADAR) project. The first is a large parallel corpus of 25 Arabic city dialects in the travel domain. The second is a lexicon of 1,045 concepts with an average of 45 words from 25 cities per concept. These resources are the first of their kind in terms of the breadth of their coverage and the fine location granularity. The focus on cities, as opposed to regions in studying Arabic dialects, opens new avenues to many areas of research from dialectology to dialect identification and machine translation.

Subfields: Artificial Intelligence, Language and Linguistics, Sociology and Political Science

View Paper: https://www.semanticscholar.org/paper/ad23a0651b0eeb87bb2bb856bea26c6b0c2f15af

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Yulia Tsvetkov
2. Sachin Kumar
3. Antonios Anastasopoulos
4. James R. Glass
5. Nizar Habash
6. Bashar Alhafni
7. Houda Bouamor
8. Ahmed M. Ali

In Language and Linguistics

1. Houda Bouamor
2. Yulia Tsvetkov
3. James R. Glass
4. Nizar Habash
5. Shuly Wintner
6. Antonios Anastasopoulos
7. Sachin Kumar
8. Bashar Alhafni

In Sociology and Political Science

1. Muhammad Abdul-Mageed
2. Yulia Tsvetkov
3. James R. Glass
4. Nizar Habash
5. Shuly Wintner

Citations in the corpus

• Machine Translation into Low-resource Language Varieties (2021, methodology): “For Arabic varieties, we use the MADAR corpus (Bouamor et al., 2018) which consists of 12K 6-way parallel sentences between English, MSA and the 4 considered varieties.”
• ADI17: A Fine-Grained Arabic Dialect Identification Dataset (2020, background): “MADAR [13, 14] (25 Arabic city dialects in the travel domain) Only text 15 (Arabic countries) -”
• ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic (2020, background): “, MSA-DA) (Elaraby and Abdul-Mageed, 2018), AOC (Zaidan and Callison-Burch, 2014) to regional (Farha and Magdy, 2020), countrylevel (Bouamor et al., 2018; Abdul-Mageed et al., 2020a; Abdelali et al., 2020), and recently province level (Abdul-Mageed et al.”
• The Interplay of Variant, Size, and Task Type in Arabic Pre-trained Language Models (2021, background): “Moreover, we collated the tweets belonging to a particular user in the MADAR Twitter Corpus in groups of 5 before feeding them to the model.”