This is a collection of past projects. I categorized them according to three topics: Master's thesis, Human-Technology Interactions, and Natural Language Processing.
In my master's thesis, I discuss bias in language model, especially in defining, measuring, and reducing bias in large language models.
-
Based on the systematic literature review, I critically analyze current methods for measuring gender bias.
-
Due to the fundamental conceptual conflation of bias, the result of existing measurement is misleading.
-
I urge the importance of theoretical soundness of defining social construct as bias, by suggesting a framework that combines philosophical and computational perspectives on bias in language models.
I designed an empirical experiment to investigate algorithmic decision making in employment, especially in gig economy.
Lastly, I conducted various NLP projects in the field of computational social science.
-
AI Hype discourse analysis
-
Extremist gardeners: Intellectual Humility and far-right Extremists in Reddit
-
Miscelleneous data science projects