Research
My research work.
Examining the Limits of Word2Vec with Toki Pona
2026 · Coauthored with Dr. Hongchen Wu
Toki Pona only has ~130 words; how effective can Word2Vec be with such a small vocabulary? We investigate model performance on a large conversational corpus of Toki Pona, developing unique evaluation methods suited to the language's small lexicon.
We find that models are effective at generating clusters of semantically similar words, whose structures mirror that of natural languages.
This research was presented at the Society for Computation in Linguistics 2026.
A Computational Approach to Analyzing Language Change and Variation in the Constructed Language Toki Pona
2025 · Coauthored with Dr. Hyoun-A Joo
Do constructed languages evolve over time like natual languages? We examine language variation in Toki Pona by developing a Toki Pona parser to analyze changes in syntactic positions of content words over time and across corpora.
Results suggest that sociolinguistic factors influence Toki Pona in the same way as natural languages; even constructed linguistic systems naturally evolve as communities use them.
This research was presented at the Emory Undergraduate Linguistics Conference 2025.
