I'm a senior studying Computer Science, Cognitive Science (BAS), and Robotics (MSE) at the University of Pennsylvania. I'm interested in developing cognitively motivated models of learning, specifically using lingustics as a framework to develop interpretable and flexible neural models. At Penn, I've worked on a variety of projects in natural language processing, computational linguistics, and psycholinguistics with Mitch Marcus, Charles Yang, and John Trueswell.
Research
Computational Linguistics (with Charles Yang)
2017 - Present
Currently, I'm making a qualitative assessment of several models of rule learning in language. Other projects I've been working on include extending Spencer Caplan's word learning model to homophone learning, and building a parameter-free sequential model of metaphor comprehension.
Paper: [upcoming] Deniz Beser and Spencer Caplan. 2019. Local Processes of Homophone Learning. SCiL 2019
Psycholinguistics (with John Trueswell)
2018 - PresentThe interaction between language and thought has been debated for decades. Following a large body of literature stating thought is independent of language unless a task implicates language, I investigate this relation in the visual domain through a series of eye-tracking experiments, and test how lemmas (abstract, conceptual word forms) are mentally activated.
Natural Language Processing (with Mitch Marcus)
2017 - 2018
I contributed to building a tool that drastically reduces Optical Character Recognition errors in corpora with minimal human labor.
Paper: Caitlin Richter, Matthew Wickes, Deniz Beser, Mitch Marcus. Low-resource Post Processing of Noisy OCR Output for Historical Corpus Digitisation. LREC 2018
I implemented and tested supervised and semi-supervised part-of-speech taggers for low-resource languages.
Poster: Part-of-Speech Tagging for Low-Resource Languages - Won first place in CIS department's poster competition!