- MSc in Computer Science, University of Minnesota, Duluth, United States
- BSc Computer Science, Ashesi University, Ghana
Teaching Statement
In teaching Computer Science in the Ghanaian context, I believe a good balance must be struck between giving students the practical skills and tools that make them immediately employable upon graduation and giving them the strong theoretical foundations needed for more advanced work in computer science. The latter is vital to ensuring that our students and graduates can develop cutting-edge ideas and technology. This, I believe, will help Ashesi’s mission of producing entrepreneurial leaders in science and engineering for Ghana and Africa. I also believe that exposure to research and researching is an important goal of undergraduate studies in Computer Science. It is my goal to expose my students to researching.
Courses taught at Ashesi
- Natural Language Processing
- Operating Systems
- Machine Learning
Research Statement
My general research area of interest is Machine Learning and, specifically, its application in speech and language processing. My immediate research objective is making it easy fordevelopers to build apps that can recognize Ghanaian languages. I think it will be pretty “awesome” to see apps that allow an illiterate farmer to request information about the going prices of their goods by simply speaking to their phone in their local language; or an app that can explain English phrases in a kid’s local language; or an app that can allow a low-literate seller to keep track of their sales or inventory through interacting with an app in their local language. Beyond being “awesome”, such technologies can help bridge the digital divide between populations literate in English and those that are not. To enable all these kinds of applications, more research is needed.
In English and other languages, referred to as high resource languages, these technologies exist and they are beginning to perform pretty well. Think of Siri or Google Assistant. Ghanaian languages, and the vast majorities of languages in the world are not high-resource languages though – they lack the scale of linguistic resources like lexicons, corpora and speech transcriptions that play an important role in existing technologies for speech and language processing. This means that novel ideas are needed to develop the technologies for them. This is a research area with high potential impact because breakthroughs in one language, say, Akan, will most likely be applicable to say Ga and, even, languages in other parts of the world. One of my primary research goals is to contribute to this. Another research goal, related to this, is studying the impact that natural language technologies could have on illiterate or low-literate people and communities. This can help highlight the importance of natural language technologies for low-resource languages and potentially encourage research funding from governments and other institutions.
Publications
- Yongle Zhang, Dennis Asamoah Owusu, Emily Gong, Shaan Chopra, Marine Carpuat, and Ge Gao. (2021) Leveraging Machine Translation to Support Distributed Teamwork Between Language-Based Subgroups: The Effects of Automated Keyword Tagging. In Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing Systems (CHI EA '21). Association for Computing Machinery, New York, NY, USA, Article 381, 1–6.
- Asamoah Owusu D. and Beaulieu J. (2018) UMDuluth-CS8761 at Semeval-2018 Task 2: Emojis: Too many choices? In Proceedings of the 12thInternational Workshop on Semantic Evaluation (New Orleans, Louisiana). Association for Computational Linguistics.
- Asamoah Owusu D. (2018) Modeling Outputs of Efficient Compressibility Estimators. Masters Thesis. University of Minnesota, Duluth.