Join

Joining as a member

Researchers, students and practitioners interested in NLP for Ethiopian languages are welcome. There is no vetting step and no fee. Members are asked to contribute to the community's work, attend events when they can, and keep to the community standards.

Add me to the community I'm registering a team Community channels All the ways to contribute

Two kinds of membership

Both categories join the same groups and events. The difference is in who decides the community's direction.

Member

Ethiopians and Ethiopian diaspora, and anyone working extensively on Ethiopian languages.

"Extensively" means a sustained body of work rather than one paper: datasets, models, supervision, or years of linguistic fieldwork. Speakers of Ethiopian languages are members whatever their field, and students count from their first project.

  • Listed in the community directory with a profile page
  • A voice in what the community prioritises and publishes
  • Can propose talks, mentor students and register a team

Affiliate member

Researchers, engineers and organisations elsewhere who work with this community or want to support it.

Collaborators on a paper, people building multilingual systems that include Ethiopian languages, and anyone who found this site and wants to help. There is no quota and no waiting period.

  • Listed in the directory, marked as an affiliate
  • The same access to the groups, seminars, mailing list and events
  • Can co-author, contribute data and code, and offer mentorship

If it is unclear which applies, choose either; a maintainer will say if the other fits better.

For groups

Registering a lab, project or student group

A team registers its Hugging Face and GitHub organisations rather than individual models. From then on, work the team publishes that covers an Ethiopian language is listed on this site automatically, and work outside those languages is not included.

Team members do not have to be listed here first.

Register a team

What we ask for

  • The team name, and a sentence on what it works on.
  • Your Hugging Face organisation, your GitHub organisation, or both.
  • A contact person, so we know who to ask.

What happens next

  • The form opens a pull request adding your team to _data/organisations.yml.
  • A maintainer reviews the entry and merges it.
  • The nightly sync reads both organisations and lists what it finds.
  • Your team appears on the ecosystem page with a live count.

Being listed

What happens after you submit the form

You supply public identifiers once; the profile is generated from them.

  1. You give us your links

    A personal site, a GitHub account, a Hugging Face account, a Google Scholar or ORCID page, whichever of those you have. One is enough; more means a fuller profile. You also pick which of the six research tracks you work on.

  2. A workflow builds your profile

    It reads those profiles, works out your name, affiliation and photo, resolves your OpenAlex author record, and opens a pull request adding a single file: _members/your-name.md.

  3. A maintainer merges it

    Usually within a day or two. You get a page at /community/your-name/.

  4. Your work is listed automatically from then on

    The nightly sync picks up your new papers from OpenAlex, Semantic Scholar and DBLP, your Ethiopian-language models and datasets from Hugging Face, and your relevant repositories from GitHub. Work you publish later is listed without a further submission.

Where the community talks

Community channels

The channels do not require you to be listed on the site first. They are covered by the community standards, and the Telegram group has rules of its own to read before posting.

Telegram

Day-to-day chat, quick questions, dataset and paper links.

Open Telegram

Discord

Topic channels, reading groups and voice meetups.

Open Discord

GitHub

Code, data pipelines and the source of this site.

Open GitHub

Hugging Face

Models and datasets, published under the EthioNLP organisation.

Open Hugging Face

Contributing

What members add

Each part of this site comes from a file in one repository. The six below are the ones members ask about most often; the contribute page lists the other ways, with the effort each takes.

Dataset or model

Datasets and models

Push it to the EthioNLP organisation on Hugging Face, or to your own account with the language tagged (am, ti, om, gez…). The nightly sync finds it and lists it on the ecosystem page. Ask in a channel for write access to the organisation.

Paper

Papers and identifiers

If your member file carries an orcid, openalex, semantic_scholar or dblp id, your papers are added automatically, including ACL Anthology entries. For records the indexes miss, add them to _data/publications_manual.yml.

Event or talk

Talks and events

The monthly seminar is for work in progress. Choose a date in the form, give a title, and a pull request is opened for you. Workshops, tutorials and hackathons use the same form.

Propose a talk
News

News items

Coverage, an award, a release, a defended thesis. Add _news/YYYY-MM-DD-slug.md with a title and a date, or send it to the mailing list and someone will add it for you.

A language with no work yet

Languages with no work yet

Many of Ethiopia's languages have no corpus, no annotation guidelines and no baseline. If you speak one, say so in a channel and members will help you scope the work.

This website

This website

Every page, script and data file is in EthioNLP/ethionlp.github.io. Corrections to a person's affiliation, a broken link or a badly worded sentence are all welcome pull requests.