Research
Commonsense Reasoning: We all share basic knowledge and reasoning ability regarding
causes and effects,
physical commonsense (if you cut the branch you sit on, you will fall),
and social commonsense (it's impolite to comment on someone's weight).
Endowing machine with such commonsense is challenging.
Such knowledge is too vast to collect from humans and often too trivial to be
mentioned in texts. In my work, I developed a model that actively
seeks additional knowledge relevant for a given situation
[1]. |
Nonmonotonic Reasoning: Everyday causal reasoning requires reasoning about the plausible but potentially defeasible conclusions from incomplete or hypothetical observations. For example, abductive reasoning ("what might explain the current events?"), counterfactual reasoning ("what if?"), and defeasible reasoning
("what might weaken my conclusion?"). I'm working on developing systems capable of nonmonotonic reasoning
for a wide range of situations describable in natural language [2, 3, 4]. |
Lexical and Compositional Semantics: Lexical variability in human language, i.e. the ability to express the same meaning in various ways, is an obstacle for natural language understanding applications. Word representations excel at capturing topical similarity (elevator/floor), as well as functional similarity (elevator/escalator), but they lack the fine-grained distinction of the specific semantic relation between a pair of words. I developed methods for recognizing lexical semantic relations between words and phrases, including ontological relationships e.g., cat is a type of animal, tail is a part of cat [5, 6]; interpreting noun-compounds, e.g. olive oil is oil made of olives while baby oil is oil for babies [7, 8]; and identifying predicate paraphrases, e.g. that X die at Y may have the same meaning as X live until Y in certain contexts [9, 10]. |
Vered Shwartz
Postdoctoral Researcher
Allen Institute for AI (AI2) and
Paul G. Allen School of Computer Science & Engineering
University of Washington
v e r e d s at allenai dot org
@VeredShwartz
Blog
CV
My research interests focus on natural language processing,
with the fundamental goal of building models capable of human-level understanding of natural language.
I'm interested in computational semantics and pragmatics, and commonsense reasoning.
I'm currently working on learning to uncover implicit meaning, which is abundant in human speech,
and on developing machines with advanced reasoning skills.
I did my PhD (2019), M.Sc. (2015), and B.Sc. (2013) in Computer Science in Bar-Ilan University. I was part of the Natural Language Processing lab where I worked on lexical and compositional semantics.
☆ I'm looking for an academic position!
Publications ▲
Filter by:
2021
Learning to Rationalize for Nonmonotonic Reasoning with Distant Supervision
Paragraph-Level Commonsense Transformers with Recurrent Memory 2020
Do Neural Language Models Overcome Reporting Bias?
Unsupervised Commonsense Question Answering with Self-Talk
"You are grounded!": Latent Name Artifacts in Pre-trained Language Models
Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning
Social Chemistry 101: Learning to Reason about Social and Moral Norms
Thinking Like a Skeptic: Defeasible Inference in Natural Language
Paraphrasing vs Coreferring: Two Sides of the Same Coin
Introductory Tutorial: Commonsense Reasoning for Natural Language Processing 2019
Still a Pain in the Neck: Evaluating Text Representations on Lexical Composition
Diversify Your Datasets: Analyzing Generalization via Controlled Variance in Adversarial Datasets
Revisiting Joint Modeling of Cross-document Entity and Event Coreference Resolution
Evaluating Text GANs as Language Models
PhD Dissertation: Learning High-Precision Lexical Inferences.
A Systematic Comparison of English Noun Compound Representations 2018
Paraphrase to Explicate: Revealing Implicit Noun-Compound Relations
Breaking NLI Systems with Sentences that Require Simple Lexical Inferences
Olive Oil Is Made of Olives, Baby Oil Is Made for Babies:
Integrating Multiplicative Features into Supervised Distributional Methods for Lexical Entailment
Neural Disambiguation of Causal Lexical Markers Based on Context
SemEval-2018 Task 9: Hypernym Discovery 2017
Acquiring Predicate Paraphrases from News Tweets
Learning Antonyms with Paraphrases and a Morphology-aware Neural Network
Hypernyms under Siege: Linguistically-motivated Artillery for Hypernymy Detection
A Consolidated Open Knowledge Representation for Multiple Texts 2016
Improving Hypernymy Detection with an Integrated Path-based and Distributional Method
Adding Context to Semantic Data-Driven Paraphrasing
Path-based vs. Distributional Information in Recognizing Lexical Semantic Relations
CogALex-V Shared Task: 2015
Learning to Exploit Structured Resources for Lexical Inference Tae-Gil Noh, Sebastian Padó, Vered Shwartz, Ido Dagan, Vivi Nastase, Kathrin Eichler, Lili Kotlerman and Meni Adler. *SEM 2015. paper bib poster short conference |
Talks ▲2021
Recent Breakthroughs and Uphill Battles in Modern Natural Language Processing 2020
Unsupervised Methods for Commonsense Reasoning
Recent Breakthroughs and Uphill Battles in Modern Natural Language Processing
What do I know? Pushing the boundaries of existing world knowledge 2019
Not a piece of cake: on lexical composition and implicit information
Fast breaking and slow building of textual inference models
What do I know? Pushing the boundaries of existing world knowledge
At Loose Ends: Challenges and Opportunities in Lexical Composition
How well can Neural Text Representations Address Lexical Composition? 2018
Multi-word Units Under the Magnifying Glass
Are we there yet? Remaining Challenges in Deep Learning based Natural Language Processing
Introduction to Natural Language Processing
Acquiring Lexical Semantic Knowledge
Paraphrase to Explicate: Revealing Implicit Noun-Compound Relations
How are you two related? Corpus-based Learning of Lexical Semantic Relations
Acquiring Lexical Semantic Knowledge 2017
Acquiring Lexical Semantic Knowledge
Acquiring Lexical Semantic Knowledge
Recognizing Lexical Inference 2016
Recognizing Lexical Inference
Recognizing Lexical Inference
Recognizing Lexical Inference 2015
Learning To Exploit Structured Resources for Lexical Inference |
Resources ▲
Probably Approximately a Scientific Blog
Introductory Tutorial on Commonsense Reasoning
Fundamentals of Deep Learning for Natural Language Processing workshop (via NVIDIA DLI)
Presentation on Crowdsourcing in NLP |