Advisors: Martha Palmer and Jim Martin
My dissertation research focuses on computational approaches to processing metaphors and other figurative language. Specifically, I'm working towards developing better representations of syntax, which has shown to be indicative of many kinds of metaphor. This is done by applying deep learning methods to syntactic structures to better represent meaning and improve classification. I hope to then use these methods to develop better formal semantic representations for non-literal language. Some preliminary work has been published at the Workshop on Figurative Language:
- Stowe, Kevin; Palmer, Martha. Leveraging Syntactic Constructions for Metaphor Identification and Interpretation. In Proceedings of the Workshop on Figurative Language Processing, held with the 56th Meeting of the Association of Computational Linguistics (ACL). pg 67-75. 2018. Melbourne, Australia.
Supervisor: Martha Palmer
I work with Professor Martha Palmer on a variety of computational lexical resources, including VerbNet and PropBank. My responsibilities include ensuring compatability with outside resources, implementation of new infrastructure, developing interface tools for other researchers, and improving the accuracy, consistency, and coverage of the resources. Our current work involves linking VerbNet with the Generative Lexicon and improving consistency among semantic predicates, and improving automatic classification with better annotation. For more, see the VerbNet website
My primary department is Linguistics. I work on NLP for figurative language, particuarly using representations of syntax to aid computational metaphor processing.
As a research intern, I did analysis of social media data (Twitter and Facebook) using machine learning algorithms, particularly clustering, to determine trends in user interactions with public company sites. We identified differences in positive and negative reactions to a variety of companies using topic modelling onsocial media, allowing for better interaction between companies and their customers
Computational BackgroundProgramming Languages
Python • Java • C++
Scikit Learn • NLTK • Keras • Gensim • PyTorch
Windows/iOS/Linux • Bash • Excel/Google Sheets • Git/SVN • XML/JSON/SQL
Linguistics BackgroundBasics Phonetics and phonology, morphology, syntax and semantics, discourse analysis, pragmatics
Dependency parsing, Combinatory Categorial Grammars (CCG), Construction Grammar (SBCG)
Predicate logic, PropBank and Abstract Meaning Representations (AMRs)