Christo Kirov

About Me

I'm currently a research software engineer at Google NYC, in the Speech and Language algorithms group. Before this, I was a Postdoctoral Fellow at the Center for Language and Speech Processing at Johns Hopkins University, where I helped start the UniMorph project. Before coming to JHU, I was a Postdoctoral Assistant Professor at the Georgetown Linguistics Department.

I received my B.A. in Computer Science and Linguistics from New York University (advisor: Adamantios Gafos), and my Ph.D. in Cognitive Science from Johns Hopkins (advisor: Colin Wilson). My current work spans multiple topics in Natural Language Processing (NLP) and Computational Linguistics. I have a special interest in NLP for low-resource, morphologically rich languages, including morphological learning, text normalization, tagging, and machine translation. I lean heavily on state-of-the-art machine learning methods, especially deep learning neural networks. I'm also interested in improving these models by focusing on system interpretability and understanding how their internal operations and representations relate to human language performance.

Details

Google Scholar

CV

Contact

ckirov gmail com

Current Activities

The UniMorph project provides morphologically-annotated lexical data and tools spanning a large, typologically diverse set of languages. Check here for an overview, or jump to our data portal here.

Initial support for UniMorph was provided by the DARPA LORELEI (Low Resource Languages for Emergent Incidents) program, focused on developing language technologies to facilitate communication in disaster-prone areas of the world where translators are scarce.

I am one of the organizers of the CoNLL-SIGMORPHON shared tasks on morphological learning across hundreds of languages. Check them out here.

Software

See my GitHub page here.

Morphological analysis software developed as part of UniMorph is available here.

Publications

Brian Roark, Christo Kirov, Cibu C Johny, Isin Demirsahin, Keith Hall, Lawrence Wolf-Sonkin, Sabrina J. Mielke (2020). Processing South Asian languages Written in the Latin Script: the Dakshina Dataset. LREC.

Amrit Nidhi, Arya McCarthy, Cassandra L. Jacobs, Christo Kirov, David Yarowsky, Ekaterina Vylomova, Garrett Nicolai, Kyle Gorman, Mans Hulden, Matteo Grella, Miikka Silfverberg, Patrick Xia, Ryan Cotterell, Sabrina J. Mielke, Yuval Pinter (2020). UniMorph 3.0: Universal Morphology. LREC.

Christo Kirov and Ryan Cotterell (2018). Recurrent Neural Networks in Linguistic Theory: Revisiting Pinker and Prince (1988) and the Past Tense Debate. TACL. arxiv

Ryan Cotterell, Christo Kirov, Mans Hulden, and Jason Eisner (2018). On the Diachronic Stability of Irregularity in Inflectional Morphology. NAACL. arxiv

Ryan Cotterell, Christo Kirov, Sabrina J. Mielke, and Jason Eisner (2018). Unsupervised Disambiguation of Syncretism in Inflected Lexicons. NAACL. New Orleans, Louisiana. pdf

Christo Kirov, Ryan Cotterell, John Sylak-Glassman, Géraldine Walther, Ekaterian Vylamova, Patrick Xia, Manaal Faruqui, Sandra Kübler, David Yarowsky, Jason Eisner, and Mans Hulden (2018). UniMorph 2.0: Universal Morphology. LREC. Miyazaki, Japan. pdf

Steven Shearing, Christo Kirov, Huda Khayrallah, and David Yarowsky (2018). Improving Low Resource Machine Translation with Morphological Glosses. AMTA. Boston, Massachusetts. pdf

Robert Churchill, Lisa Singh, and Christo Kirov (2018). A Temporal Topic Model for Noisy Mediums. PAKDD. Melbourne, Australia.

Ryan Cotterell, Christo Kirov, Mans Hulden, and Jason Eisner (2017). On the Complexity and Typology of Inflectional Morphological Systems. TACL.

Ryan Cotterell, Ekaterina Vylamova, Huda Khayrallah, Christo Kirov, and David Yarowsky (2017). Paradigm Completion for Derivational Morphology. EMNLP. Copenhagen, Denmark. pdf

Ryan Cotterell, Christo Kirov, John Sylak-Glassman, Géraldine Walther, Ekaterina Vylamova, Patrick Xia, Manaal Faruqui, Sandra Kübler, David Yarowsky, Jason Eisner, and Mans Hulden (2017). CoNLL-SIGMORPHON 2017 Shared Task: Universal Morphological Reinflection in 52 Languages. CoNLL. Vancouver, Canada. pdf

Christo Kirov, John Sylak-Glassman, Ryan Cotterell, Rebecca Knowles, and Matt Post (2017). A Rich Morphological Tagger for English: Exploring the Cross-Linguistic Tradeoff Between Morphology and Syntax. EACL. Valencia, Spain. pdf

Ryan Cotterell, Christo Kirov, and John Sylak-Glassman (2017). Neural Graphical Models over Strings for Principal Parts Morphological Paradigm Completion. EACL. Valencia, Spain. Outstanding Paper Award. pdf

Ryan Cotterell, Christo Kirov, John Sylak-Glassman, David Yarowsky, Jason Eisner, and Mans Hulden (2016). The SIGMORPHON 2016 Shared Task - Morphological Reinflection. Proceedings of the 2016 Meeting of SIGMORPHON. ACL. Linguistics. Berlin, Germany. pdf

Christo Kirov, John Sylak-Glassman, Roger Que, and David Yarowsky (2016). Very-large Scale Parsing and Normalization of Wiktionary Morphological Paradigms. LREC. Portoroz, Slovenia. pdf

John Sylak-Glassman, Christo Kirov, and David Yarowsky (2016).Remote Elicitation of Inflectional Paradigms to Seed Morphological Analysis in Low-Resource Languages. LREC. Portoroz, Slovenia. pdf

John Sylak-Glassman, Christo Kirov, David Yarowsky, and Roger Que (2015). A Language-Independent Feature Schema for Inflectional Morphology. ACL. Beijing, China. pdf

John Sylak-Glassman, Christo Kirov, David Yarowsky, and Roger Que (2015). A Universal Schema for Rich Morphological Annotation. SFCM. Stuttgart, Germany. pdf

Christo Kirov and Colin Wilson (2013). Bayesian Speech Production: Evidence from Latency and Hyperarticulation. COGSCI. Berlin, Germany. pdf

Christo Kirov and Colin Wilson (2012). The Specificity of Online Variation in Speech Production. COGSCI. Sapporo, Japan. pdf

Christo Kirov and Bob Frank (2011). Processing of Nested and Cross-Serial Dependencies: an Automaton Perspective on SRN Behavior. Connection Science. Volume 24. Issue 1. pp. 1-24. pdf

Christo Kirov and Adamantios Gafos (2010). Assembling Phonological Representations. Phonological and Complex Adaptive Systems: Phonology and Complexity. Chitoran, Coupe, Marsico, Pellegrino (eds.), Mouton de Gruyter, Berlin/New York.

Christo Kirov and Adamantios Gafos (2007). Dynamic Phonetic Detail in Lexical Representations. ICPHS. Saarbrucken, Germany. pp. 637-640.

Invited Talks

Christo Kirov (2018). Computational Morphology with UniMorph. Army Research Laboratories

Christo Kirov (2017). Computational Morphophonology - Bridging NLP and Linguistics. University of Chicago.

Christo Kirov (2017). Morphological Modeling for Multi-lingual NLP. GUCL. Georgetown University.

Christo Kirov (2016). Building Strong Inductive Baselines for Computational Morphophonology. New York University.

Christo Kirov (2016). UniMorph - Universal Morphology at JHU. Center For Language and Speech Processing Seminar Series. Johns Hopkins University. video

Christo Kirov (2015). The Consequences of Bayesian Decision-Making in Speech Production. Common Ground Speaker Series. University of Pennsylvania.

Presentations

Ryan Cotterell, Christo Kirov, Mans Hulden, and Jason Eisner (2018). The Pareto Complexity of Inflectional Systems. SCIL. Salt Lake City, Utah. slides

Christo Kirov (2017). Recurrent Neural Networks as a Strong Baseline for Morphophonological Learning. LSA. Austin, TX. poster

John Sylak-Glassman, Christo Kirov, David Yarowsky, and Roger Que (2015). Typology Enriching NLP: A Universal Schema for Inflectional Morphology. ALT. New Mexico.

Christo Kirov (2014). Grammatical Influences in a Bayesian Speech Production Framework. COGSCI. Quebec City, Quebec. poster

Christo Kirov and Colin Wilson (2013). Modeling the Relationship Between Competition, Latency, and Articulation. LSA. Boston, MA.

Christo Kirov and Colin Wilson (2012). How Phonological Context Affects Phonetic Realization. LABPHON. Stuttgart, Germany.

Christo Kirov and Colin Wilson (2012). Modeling Speech Production with Bayesian Inference: Competition, Latency, and Articulation. APCAM. Minneapolis, MN.

Christo Kirov and Colin Wilson (2012). Specificity of Online Variation in Speech Production. LSA, Portland, Oregon.

Christo Kirov and Colin Wilson (2011). Explaining Online Hyperarticulation in Speech Production. NECPHON, Yale University, New Haven, Connecticut.

Christo Kirov (2008). Forward and Reverse Recall in Simple Recurrent Networks. COGSCI, Washington, D.C.

Teaching

Natural Language Processing (Georgetown University, Fall 2014)

Phonology/Phonetics I (Johns Hopkins University, Spring 2014)

Computational Phonology (Georgetown University, Spring 2014)

Computational Corpus Linguistics (Georgetown University, Fall 2013)

Computational Corpus Linguistics (Georgetown University, Fall 2012)