📧 hellina underscore nigatu at berkeley dot edu
I am a PhD student at the University of California, Berkeley in the EECS department. I finished my undergrad at Addis Ababa University, Ethiopia in Electrical and Computer Engineering and received my Masters of Science in CS from UC Berkeley in 2023. I am currently co-advised by Prof. Sarah Chasins and Prof. John Canny. I am part of PLAIT, CannyLab and BAIR.
My research interest broadly lies in the intersection of AI and HCI, with a focus on making usable AI tools accessible to end users. I am currently looking into making NLP tools usable and accessible for low-resourced languages. I am also interested in the impact of AI on society, specifically in how it affects Global Majority countries and communities.
Outside of research, I like to read books📚, make and drink traditional Ethiopian coffee☕, knit🧣, explore art🎨, and write✍️.
Presented poster on our Online Harm FAccT’24 paper at the Stanford-Berkeley Research Meetup.
Paper on online harm due to policy-violating sexual content on YouTube in Amharic Accepted to FAccT 2024
Paper on need-finding work with low-resourced language contributors and their interaction with online knowledge repository accepted to CHI 2024
I received the ACM-W scholarship to attend EMNLP’23 where I will be presenting my work. Thank you to the ACM-W scholarship and Google for sponsoring my travel.
LLM evaluation work [poster] done with my mentee Bontu Fufa Balcha from my course accepted to two workshops!! Black in AI workshop co-located with NeurIPS’23 and WiNLP workshop co-located with EMNLP’23.
I built a library…not the code type…a physical library 🏫…where you find books :-). Read more about it here.
I will be TAing for Addis Coder in Summer 2023!!
Finally translated my website content to Amharic. More languages to come :-))
Was an invited speaker at the Wade Scholarship Program welcome event for 2022 Scholars.
Delighted to announce I am one of the 2022 winners of the SIGHPC Computational and Data Science Fellowship.
Excited to be featured with other Ethiopian women in STEM on BBC Amharic.
[Link Upcoming] Hellina Hailu Nigatu, Inioluwa Deborah Raji. ““I Searched for a Religious Song in Amharic and Got Sexual Content Instead”: Investigating Online Harm in Low-Resourced Languages on YouTube.” Proceedings of ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT)–Forthcoming.
[paper] Hellina Hailu Nigatu, John Canny, Sarah Chasins. “Low-Resourced Languages and Online Knowledge Repositories: A Need-Finding Study” Proceedings of ACM Conference on Human Factors in Computing Systems (ACM CHI)–Forthcoming.
[paper] Hellina Hailu Nigatu*, Atnafu Lambebo Tonja*, Jugal Kalita. “The Less the Merrier? Investigating Language Representation in Multilingual Models” Proceedings of Empirical Natural Language Processing (EMNLP) 2023 Findings.
[paper] Hellina Hailu Nigatu, Lisa Pickoff-White, John Canny, Sarah Chasins. “Co-Designing for Transparency: Lessons from Building a Document Organization Tool for the Criminal Justice Domain.” Proceedings of ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT).
[paper] Hellina Hailu Nigatu, John Canny, Sarah Chasins. “A Need Finding Study with Low-Resourced Language Content Creators” Proceedings of 4th ACM African Human-Computer Interaction Conference (AfriCHI 2023)
[paper] Atnafu Lambebo Tonja, Hellina Hailu Nigatu, Olga Kolesnikova, Grigori Sidorov, Alexander Gelbukh, Jugal Kalita. “Enhancing Translation for Indigenous Languages: Experiments with Multilingual Models.” For AmericasNLP 2023 Shared Task on Machine Translation into Indigenous Languages.
[paper] Andrew McCrabb, Hellina Nigatu, Absalat Getachew, Valeria Bertacco. “DyGraph: a dynamic graph generator and benchmark suite.” Proceedings of the 5th ACM SIGMOD Joint International Workshop on Graph Data Management Experiences & Systems (GRADES) and Network Data Analytics (NDA).
[paper] Chris Chinenye Emezue, Hellina Hailu Nigatu, Cynthia Thinwa, Lerato Louis, Idris Abdulmumin, Samuel Gbenga Oyerinde, Benjamin Ayoade Ajibade, Helper Zhou, Emeka Felix Onwuegbuzia, Handel Chiagozie Emezue, Ifeoluwatayo Adeseye Ige, Atnafu Lambebo Tonja, Chiamaka Ijeoma Chukwuneke, Shamsuddeen Hassan Muhammad, Olanrewaju Samuel. “The African Stopwords Project: Curating Stopwords for African Languages” 3rd Workshop on African Natural Language Processing.
[video] Starting with the Need: How we can use HCI to inform technological design for low-resourced languages. Speaker for IoTDayWomen2024. Apr, 2024. Virtual
DOT: Reflections from Building a Document Organisation Tool in the Criminal Justice Domain Invited Speaker for Datathon4Jusitce 2023. Nov, 2023. Virtual
Challenges and Prospects: Building a Document Organisation Tool for Processing Police Misconduct Data Invited Anchor Speaker for Ingram Olkin Forum on Police Use of Force. Nov, 2023. Northfield, MN
Auditing YouTube Recommendation in Low-Resourced Languages: A Case Study in Amharic. The 7th Workshop on Online Abuse and Harms. July, 2023.
Human-Centered and Ethical NLP for Low-Resourced Languages. Invited talk for Learning Session at WhoseKnowledge?. July, 2023.
Build Your Case: Using AI and HCI to aid Document Organization Invited talk at Stanford for Watchdog Reporting Class April 2022 Stanford, CA
Detecting and Exploring Replicated Files in Large Document Dumps: A demonstration workshop for investigative journalists on using document organizing tool I am building. April 2022 Stanford, CA