Why I'm doing this

In the final months of 2021, in search of a final year undergraduate thesis topic, especially one that would enable me to make a foray into the area of Natural Language Processing, I stumbled upon the concept of code-switching while reading Dan Jurafsky’s Speech and Language processing.

Code-switching involves mixing two or more languages within an expression. At the simplest level, code-switching is what happens when a Nigerian says “bawo ni, what will you eat?", an expression that combines the Yoruba and English languages and means “how are you, what will you eat” in plain English. Code-switched expressions suffer from more information loss during translation than expressions in a single language.

After some research, I found that there was little to no existing research in the area of code-switched Yoruba-English. As such, I’ve decided to focus on researching how code-switched Yoruba-English expressions can be augmented to reduce information loss. This will happen by employing methods such as context analysis to reduce the code-switched expression to a single-language expression. If you think it sounds like challenging work, it’s safe to say you’re correct, but what’s more significant is the fact that it’s as interesting as it’s challenging.

Over the next few months, I’m going to experiment the learning by sharing model to help me get up to speed with the knowledge I need for my research.

From time to time, I might also write about my experience learning other things revolving around my day to day work as a backend engineer, and my experience learning how to build apps using Solidity, an area in which I’m finding growing interest.