Jacob Collard

Cornell University

Department of Linguistics

jacob@thorsonlinguistics.com

About

I am a Ph.D. student in the Cornell University department of linguistics and a consultant at the National Institute of Standards and Technology. My main research interests are in minimally supervised learning models for syntax and semantics and computational models of semantics and pragmatics. I am a member of the Computational Linguistics Lab and the NLP group.

Research

I am currently working on a model of possible world semantics capable of efficiently representing regular sets of worlds and relations between them. The system is fully compositional, with well-defined lexical semantics. Though the system is somewhat limited in its expressive power, it is capable of representing many semantic operations, including definiteness, modality, plurality, and presupposition. The system can efficiently be used for applications such as natural language interfaces for games or knowledge systems, and provides insight into the cognitive science of language-based reasoning.

I am also working on an algorithm for unsupervised learning of syntax in various formalisms, including Combinatory Categorial Grammars, Relational Grammars, and Minimalist Grammars. This algorithm learns purely from strings, using probabilistic heuristics to improve learnability.

Teaching

  • COGST1101: Introduction to Cognitive Science (Spring 2018) (TA)
  • LING1101: Introduction to Linguistics (Fall 2017) (TA)
  • LING4424: Computational Linguistics (Spring 2017) (TA)
  • LING1100-103: The Death of Language (Fall 2016) (Instructor)

Service

I have worked on the editing team for Semantics and Linguistic Theory 26 and 27 and I am currently on the organizing committe for NELS 49. I have also helped run several workshops for the Cornell Linguistics Circle on topics including natural language processing, LaTeX for linguists, and basic web development.