Скачать или смотреть Hate speech detection: Bias in data and annotations | Sandra Kübler | DiLCo Video Reader

Hate speech detection: Bias in data and annotations

Sandra Kuebler | Indiana University Bloomington | 2024

Hate speech detection is popular topic with important real world applications. In this talk, I want to have a closer look at the data situation in hate speech detection. Using machine learning for hate speech detection tends to give good results, but such results are often deceptive. Issues that that have been raised are biases in the data, annotation quality, sampling strategies, domain effects, but also include the definition of what we consider hate speech. All of these issues have a profound effect on results, but do not have simple solutions.

Sandra Kübler is a Professor of Computational Linguistics in the Department of Linguistics at Indiana University. Her main research areas are syntactic parsing, with a focus on morphologically rich languages, and machine learning approaches to sentiment analysis, emotion detection, hate speech detection, and conspiracy detection, with a focus on decisions made during data collection and annotation. Dr. Kübler currently serves as an associate editor for the journal Natural Language Engineering. Academic webpage: https://cl.indiana.edu/~skuebler/

This talk was originally presented on 28.03.2024 as part of DiLCo Lecture Series 2024.

DiLCo (‘Digital language variation in context’) was an international research network at Universität Hamburg, funded by the Excellence Strategy of the German federal and state governments (https://www.dilco.uni-hamburg.de).