I developed rigorous evaluations for machine reading comprehension systems. Some of the resulting ideas were written up in a 2020 research paper.
I also implemented extensive logic rules for a travel-oriented demonstration application of the company's automated reasoning tools.
During my two internships at Google
I developed models for rating entities' centrality within a document. Published in a 2014 EACL paper.
I also explored techniques for identifying high-quality responses to controversial Internet articles.
As a Ph.D. student, I studied methods for extracting structured semantic representations from natural-language text. I looked for ways to incorporate linguistic insights into natural language technologies, most notably the principles of Construction Grammar.
I developed an annotation scheme and associated corpus called BECAUSE, which annotated the cause-and-effect relations stated by a text. The scheme was designed to represent causal relations expressed by nearly any linguistic construction, not just discrete words and phrases.
I also developed two systems for automatically extracting and classifying causal relations: Causeway, a tagger involving learned lexico-syntactic patterns and feature-engineered classifiers, and DeepCx, a neural transition-based tagger.
Check Google Scholar for a complete and up-to-date list of my publications.
- Dunietz, Jesse, Gregory Burnham, Akash Bharadwaj, Jennifer Chu-Carroll, Owen Rambow, David Ferrucci." To Test Machine Comprehension, Start by Defining Comprehension." ACL 2020.
- Dunietz, Jesse, Lori Levin, and Jaime Carbonell. "DeepCx: A transition-based approach for shallow semantic parsing with complex constructional triggers." EMNLP 2018.
- Thesis: Dunietz, Jesse. Annotating and Automatically Tagging Constructions of Causal Language (2018). Ph.D. thesis, Carnegie Mellon University, Pittsburgh, PA.
- Dunietz, Jesse, Lori Levin, and Jaime Carbonell. "The BECauSE Corpus 2.0: Annotating Causality and Overlapping Relations." LAW XI – The 11th Linguistic Annotation Workshop (2017).
- Dunietz, Jesse, Lori Levin, and Jaime Carbonell. "Automatically Tagging Constructions of Causation and Their Slot-Fillers." Transactions of the Association for Computational Linguistics (2017).
- Dunietz, Jesse, Lori Levin, and Jaime Carbonell. "Annotating Causal Language Using Corpus Lexicography of Constructions." Proceedings of LAW IX – The 9th Linguistic Annotation Workshop (2015).
- Dunietz, Jesse, and Dan Gillick. "A New Entity Salience Task with Millions of Training Examples." EACL 2014.
- Dunietz, Jesse, Lori Levin, and Jaime Carbonell. "The Effects of Lexical Resource Quality on Preference Violation Detection." ACL 2013.
- Dunietz, Jesse, Lori Levin, and Miriam R. L. Petruck. "Construction Detection in a Conventional NLP Pipeline." AAAI Spring Symposium Technical Report SS-17-02: Computational Construction Grammar and Natural Language Understanding (2017).
- Dunietz, Jesse. "PyDecay/GraphPhys: A Unified Language and Storage System for Particle Decay Process Descriptions." Accepted for publication in DOE Journal of Undergraduate Research, Vol. XI (canceled for funding reasons). Presented as a student poster at the 2011 AAAS Annual Meeting.