Maria Ryskina

I am a PhD student at the Language Technologies Institute of the School of Computer Science at Carnegie Mellon University. I am fortunate to be advised by Matt Gormley, Eduard Hovy, and Taylor Berg-Kirkpatrick.

I work on machine learning models for computational linguistics, and I’m especially excited about idiosyncratic uses of language and how they lead to larger-scale language change.

I recieved my bachelor’s and master’s degrees from the Moscow Institute of Physics and Technology and Skoltech. I’ve worked on my master’s thesis at MIT InfoLab and interned at DiDi Labs with Kevin Knight.

news

Jun 27, 2021	I am a Diverse Intelligences Summer Institute 2021 Fellow! I’ll be spending the next three weeks learning about other disciplines’ perspectives on intelligence.
Jun 20, 2021	My lecture on informal romanization is now on YouTube! I am also giving a shorter version of this talk at NLPhD series @ Saarland University on June 22 at 14:30 CEST.
Jun 14, 2021	On June 18 at 17:00 UTC, I’ll be giving a talk about my work on informal romanization as a part of SIGTYP summer lecture series. Register to attend here!
Apr 7, 2021	We will be presenting our work on NoiseQA at EACL 2021! Come find Abhilasha Ravichander, Siddharth Dalmia and me on April 21 14:00-17:00 CEST!
Mar 8, 2021	I am helping organize SIGMORPHON 2021 Shared Task 0: Generalization in Morphological Inflection Generation. Register to participate here!

selected publications

ACL

Phonetic and Visual Priors for Decipherment of Informal Romanization

Maria Ryskina, Matthew R. Gormley, and Taylor Berg-Kirkpatrick

In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics 2020

Abstract arXiv Code Slides Talk

Informal romanization is an idiosyncratic process used by humans in informal digital communication to encode non-Latin script languages into Latin character sets found on common keyboards. Character substitution choices differ between users but have been shown to be governed by the same main principles observed across a variety of languages—namely, character pairs are often associated through phonetic or visual similarity. We propose a noisy-channel WFST cascade model for deciphering the original non-Latin script from observed romanized text in an unsupervised fashion. We train our model directly on romanized data from two languages: Egyptian Arabic and Russian. We demonstrate that adding inductive bias through phonetic and visual priors on character mappings substantially improves the model’s performance on both languages, yielding results much closer to the supervised skyline. Finally, we introduce a new dataset of romanized Russian, collected from a Russian social network website and partially annotated for our experiments.
SCiL

Where New Words Are Born: Distributional Semantic Analysis of Neologisms and Their Semantic Neighborhoods

Maria Ryskina, Ella Rabinovich, Taylor Berg-Kirkpatrick, David R. Mortensen, and Yulia Tsvetkov

In Proceedings of the Society for Computation in Linguistics 2020

Abstract arXiv Code Poster

We perform statistical analysis of the phenomenon of neology, the process by which new words emerge in a language, using large diachronic corpora of English. We investigate the importance of two factors, semantic sparsity and frequency growth rates of semantic neighbors, formalized in the distributional semantics paradigm. We show that both factors are predictive of word emergence although we find more support for the latter hypothesis. Besides presenting a new linguistic application of distributional semantics, this study tackles the linguistic question of the role of language-internal factors (in our case, sparsity) in language change motivated by language-external factors (reflected in frequency growth).
ACL

Automatic Compositor Attribution in the First Folio of Shakespeare

Maria Ryskina, Hannah Alpert-Abrams, Dan Garrette, and Taylor Berg-Kirkpatrick

In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics 2017

Abstract arXiv Poster

Compositor attribution, the clustering of pages in a historical printed document by the individual who set the type, is a bibliographic task that relies on analysis of orthographic variation and inspection of visual details of the printed page. In this paper, we introduce a novel unsupervised model that jointly describes the textual and visual features needed to distinguish compositors. Applied to images of Shakespeare’s First Folio, our model predicts attributions that agree with the manual judgements of bibliographers with an accuracy of 87%, even on text that is the output of OCR.