Link Search Menu Expand Document

Pidgin Wiki

Pidgin.wiki is a free online reference for West African Pidgin English spoken in different parts of Africa including Cameroon, Nigeria, Ghana and Seria Leone. The aim of the Pidgin.wiki project is to;

  • Build a corpus for Pidgin English.
  • Provide Pidgin word definations and meanings.
  • Understand patterns in Pidgin hence proposing a standard structure.
  • Provide a machine model/API to translate from and to Pidgin English.
  • Promote the use of Pidgin English, as blocker to barriers in Education.

Background

4 West African countries speak Pidgin English. In Nigeria alone more than 75 million people communicate in Pidgin [ref]. Also, Southern Cameroon comprising the English-speaking regions of Cameroon accounts for more than 5 million fluent speakers with at least 2 million being native speakers [ref], in addition about 50% of Cameroonians can speak some Pidgin taking the toll up to over 10 million speakers in Cameroon. Ghana is known to also have at least 5 million native speakers [ref]. When Equatorial Guinea, Seria Leone and countries influenced by these combined culture are taken into consideration we easily have more than 100 million africans who can speak and understand Pidgin English in some form.

Country Speakers
Cameroon 10 million
Ghana 5 million
Nigeria 75 million

Table : Top Speakers

With so many people speaking Pidgin English why don’t computers understand it?

  • Why do the very best translation algorithms that exist fail to translate Pidgin English?
  • Why is Education not delivered in Pidgin English?
  • Why is it even outlawed in certain institutions within speaking nations?

We could not find logical answers to these questions, at least ones that justify the status quo. We absolutely think Pidgin English should be way more applicable and we have set out to make that happen.

Prior development efforts

Before this project, there have been some other work done to achieve some of the goals outlined here with varying amount of success, mostly less than 1% for most work reviewed. See;

  • http://naijalingo.com/ (Nigerian Pidgin English Dictionary)
  • https://www.bbc.com/pidgin (BBC News served in Pidgin English)

As an open source organization we are building different tools that can help us achieve our ambitious dreams for Pidgin English. These tools include but are not limited to;

  • Pigdin.Wiki (Free cyclopedia displaying the combined results of reference material available in Pidgin English)
  • Bandolo (A data collection and management bot)
  • Langwa (NMT Core)

See Architecture