I am a data scientist and linguist in Chicago.
Coding projects
I created an open source Python package to generate realistic synthetic data that protects the privacy of people in the training data. Slides from my talk at PyData NYC 2019 can be found here.
Python code to transcribe English and Finnish text into the International Phonetic Alphabet (IPA), as well as some tools for phonological features.
I made a Javascript visualization of the movement of CTA "L" trains over the course of an hour.
I made a Javascript dictionary for looking up words in Ancient North Arabian.
Publications and manuscripts
Using variation in oxygen isotopes from tooth enamel, it is possible to estimate what season herd animals were born. I contributed to the statistical analysis of this type of archeological data.
I built unsupervised statistical models that learn to identify phrase boundaries in American Sign Language based on prosodic cues.
I helped study the acquisition of prosodic cues to mark phrase boundaries in American Sign Language. We found that children as young as five years old already show a similar distribution of cues to adult signers.
Imperatives can be used to perform a variety of speech acts, including commands, permission, and advice. I helped study what prosodic cues distinguish these different speech acts in American Sign Language, as well as how interpretable the cues are.
I performed a corpus study of sibilant harmony in Basque. I found patterns that cannot be explained under a standard Agreement By Correspondence account of harmony, and offered a modification to the theory that captures the data.
I investigated Shota Rustaveli's use of metrical structure in The Knight in the Panther's Skin, often considered the national epic of Georgia. I found patterns in the placement of word boundaries that help mark the ends of lines and stanzas.
Other activities
I am a contributing editor at the South Side Weekly, an alt weekly based in Chicago.
I play accordion and oboe in the University of Chicago Middle East Music Ensemble.