Skip to content

Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions.

Notifications You must be signed in to change notification settings

SaiedAlshahrani/Analysis-of-Arabic-Wikipedias

 
 

Repository files navigation

Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions

In this repository, we contribute the scripts and data we used to analyze Arabic Wikipedia editions: Modern Standard Arabic, Egyptian Arabic, and Moroccan Arabic, comparing them to the English Wikipedia using Wikimedia Statistics (Wikistats) API.

ACL Citation:

Saied Alshahrani, Esma Wali, and Jeanna Matthews. 2022. Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions. In Proceedings of the The Seventh Arabic Natural Language Processing Workshop (WANLP), pages 361–371, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.

Citation BibTeX:

@inproceedings{alshahrani-etal-2022-learning,
    title = "Learning From {A}rabic Corpora But Not Always From {A}rabic Speakers: A Case Study of the {A}rabic {W}ikipedia Editions",
    author = "Alshahrani, Saied  and
      Wali, Esma  and
      Matthews, Jeanna",
    booktitle = "Proceedings of the The Seventh Arabic Natural Language Processing Workshop (WANLP)",
    month = dec,
    year = "2022",
    address = "Abu Dhabi, United Arab Emirates (Hybrid)",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.wanlp-1.34",
    pages = "361--371",
    abstract = "Wikipedia is a common source of training data for Natural Language Processing (NLP) research, especially as a source for corpora in languages other than English. However, for many downstream NLP tasks, it is important to understand the degree to which these corpora reflect representative contributions of native speakers. In particular, many entries in a given language may be translated from other languages or produced through other automated mechanisms. Language models built using corpora like Wikipedia can embed history, culture, bias, stereotypes, politics, and more, but it is important to understand whose views are actually being represented. In this paper, we present a case study focusing specifically on differences among the Arabic Wikipedia editions (Modern Standard Arabic, Egyptian, and Moroccan). In particular, we document issues in the Egyptian Arabic Wikipedia with automatic creation/generation and translation of content pages from English without human supervision. These issues could substantially affect the performance and accuracy of Large Language Models (LLMs) trained from these corpora, producing models that lack the cultural richness and meaningful representation of native speakers. Fortunately, the metadata maintained by Wikipedia provides visibility into these issues, but unfortunately, this is not the case for all corpora used to train LLMs.",
}

About

Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.2%
  • Python 0.8%