1

Vicarious Offense and Noise Audit of Offensive Speech Classifiers: Unifying Human and Machine Disagreement on What is Offensive
Offensive speech detection is a key component of content moderation. However, what is offensive can be highly subjective. This paper …
PDF Cite Dataset
Vicarious Offense and Noise Audit of Offensive Speech Classifiers: Unifying Human and Machine Disagreement on What is Offensive
Disagreement Matters: Preserving Label Diversity by Jointly Modeling Item and Annotator Label Distributions with DisCo
Annotator disagreement is common whenever human judgment is needed for supervised learning. It is conventional to assume that one label …
PDF Cite
Subjective Crowd Disagreements for Subjective Data: Uncovering Meaningful CrowdOpinion with Population-level Learning
Human-annotated data plays a critical role in the fairness of AI systems, including those that deal with life-altering decisions or …
PDF Cite
Findings from the Bambara - French Machine Translation Competition (BFMT 2023)
Orange Silicon Valley hosted a low-resource machine translation (MT) competition with monetary prizes. The goals of the competition …
PDF Cite
Annotator Response Distributions as a Sampling Frame
Annotator disagreement is often dismissed as noise or the result of poor annotation process quality. Others have argued that it can be …
PDF Cite
Improving Label Quality by Joint Probabilistic Modeling of Items and Annotators
We propose a fully Bayesian framework for learning ground truth labels from noisy annotators. Our framework ensures scalability by …
PDF Cite
A framework for automated corpus compilation for KeyXtract: Twitter model
The corpus is a limiting factor for a keyword extraction process with a word matching stage. This paper proposes a framework to …
PDF Cite DOI
A method to extract essential keywords from a tweet using NLP tools
A tweet is an authentic use of Natural Language where the user has to deliver the message in 140 characters or less. According to …
PDF Cite DOI
KeyXtract Twitter Model - An Essential Keywords Extraction Model for Twitter Designed using NLP Tools
Since a tweet is limited to 140 characters, it is ambiguous and difficult for traditional Natural Language Processing (NLP) tools to …
PDF Cite