Program

The workshop format has been updated to try be as inclusive of as many timezones as possible. The workshop will be split into a UTC Morning session which is Europe, Africa and Asia friendly, and a UTC Evening session which is more America friendly.

Virtual Logistics

Morning Session (UTC)

Morning Invited & Contributed Talks (UTC)

Time Title Speaker Resources  
7:00 AM Opening Remarks Kathleen Siminyu    
7:10 AM Invited Talk: Masakhane, A Focus on Machine Translation for African Languages Jade Abbott video paper, slides  
7:45 AM Contributed Talk: Towards Supervised and Unsupervised Neural Machine Translation Baselines for Nigerian Pidgin Orevaoghene Ahia video paper  
8:05 AM Contributed Talk: Distant Supervision and Noisy Label Learning for Low Resource Named Entity Recognition: A Study on Hausa and Yorùbá David Adelani video paper, poster  
8:25 AM AI4D Challenge Winner: Crowdsourcing a Wolof ASR dataset using a custom built tool Thierno Diop video slides  

Morning Poster Session (UTC)

Time Title Speaker Resources
8:35 AM Tunizi: A Tunisian Arabizi Sentiment Analysis Dataset Chayma Fourati video poster
8:40 AM The Modified Segmentation Algorithm for Recognition of Ancient Ge’ez Scripts Written on the Vellum Girma Sisay  
8:45 AM Unsupervised Pidgin Text Generation By Pivoting English Data and Self-Training Ernie Chang videos paper, poster
8:50 AM Using LSTM Networks to Translate from French to Senegalese Local Languages: Wolof as a Case Study Alla Lo videos paper, poster
8:55 AM Amharic Abstractive Text Summarization Amr Zaki videos paper, poster
9:00 AM A Hierarchical Transformer for Unsupervised Parsing Ashok Thillaisundaram video paper, poster
9:05 AM On optimal transformer depth for low-resource language translation Elan van Biljon video paper, poster
9:10 AM Combining Pretrained High-Resource Embeddings and Subword Representations for Low-Resource Languages Machel Reid video paper, poster
9:15 AM Evaluating Amharic Machine Translation Asmelash Teka Hadgu video paper, poster
9:20 AM FFR V1.0: Fon-French Neural Machine Translation Chris Emezue & Bonaventure Dossou video paper, poster
9:25 AM Using Punkt for Sentence Segmentation in non-Latin Scripts: Experiments on Kurdish (Sorani) Texts Roshna Omer Abdulrahman video poster
9:30 AM Tigrinya Neural Machine Translation with Transfer Learning for Humanitarian Response Alp Oktem video paper, poster
9:45 AM Invited Talk: IntrodUTCion to Multilingual Wikipedia Data for NLP Isaac Johnson video slides

Morning Q&A with Invited & Contributed talk speakers (UTC)

10:00 AM LIVE Q&A with Invited & Contributed talk speakers Jade Abbott, Orevaoghene Ahia, David Adelani, Theirno Diop, Isaac Johnson, moderated by Kathleen Siminyu  

Midday Datathon (UTC)

The Wikimedia team are hosting a Datathon which will run from 10:30AM - 15:00, in between the morning and evening session. So bring along your grandmother, your partner, your roommate, and let’s work together to contribute to wikipedia for our native languages!

Here is the talk about what you can do

Here are some useful links for when you need them:

Isaac Johnson from Wikimedia is available on Rocket chat to help you start translating data for your language!

While you’re contributing, feel free to checkout the presentations by AI4Dev projects, across Africa.

Evening Session (UTC)

Evening Invited & Contributed Talks (UTC)

Time Title Speaker Resources  
15:00 Invited Talk: The Low-resource Natural Language Processing Toolbox, 2020 Version Graham Neubig slides  
15:35 Panel Discussion: What are African NLP’s Moonshot Problems? Natalie Schluter, Sebastian Ruder, Surafel Melaku Lakew, moderated by Jade Abbott    
16:10 Contributed Talk: Towards A Sign Language Gloss Representation Of Modern Standard Arabic Salma El Anigri poster  
16:30 Contributed Talk: Improving Yorùbá Diacritic Restoration David Adelani paper, poster  
16:50 AI4D Challenge Winner: Creating MT and ASR datasets from online sources (video may not be available) Dosseh Ameck Guy Max Desire slides  

Evening Poster Session (UTC)

Time Title Speaker Resources
17:00 Investigating an approach for low resource language dataset creation, curation and classification: Setswana and Sepedi Vukosi Marivate paper, slides
17:05 Pooled Embeddings for Unsupervised IndUTCion of Yorùbá Morphology Olamilekan Wahab  
17:10 Neural Machine Translation for South Africa’s Official Languages (video wont be available) Jason Webster  
17:15 Assessing Human Translations from French to Bambara for Machine Learning: a Pilot Study Allahsera Auguste Tapo paper, poster
17:20 A Benchmark Dataset for Igbo-English Machine Translation Ignatius Ezeani paper, poster
17:25 Bert Fine-tuning For Arabic Text Summarization Khalid Elmadani poster
17:30 Semantic Enrichment of Nigerian Pidgin English for Contextual Sentiment Classification Wuraola Oyewusi paper, poster
17:35 Low Resource Neural Machine Translation: A Benchmark for Five African Languages Surafel Melaku Lakew paper, poster
17:40 Neurals Networks for Projecting Named Entities from English to Ewondo Michael Franklin Mbouopda poster
17:45 Neural Machine Translation for African Languages Xian Li poster
17:50 Towards Neural Machine Translation for Edoid Languages Iroro Orife paper, poster

Evening Q&A with Invited & Contributed talk speakers

18:00 LIVE Q&A with Invited & Contributed talk Speakers Graham Neubig, Salma El Anigri, Iroro Orife, David Adelani Dosseh Ameck Guy Max Desire  
18:30 Closing Remarks Vukosi Marivate