The LITHME WG1 on Computational Linguistics is organizing an invited talk session with Valerio Basile (University of Turin), to take place online via Zoom on February 14, 2022, at 6 pm CET.
Title: The Perspectivist Approach to Natural Language Processing
Abstract:
Much of contemporary Artificial Intelligence is based on supervised machine learning, a methodology that leverages large manually annotated datasets. The harmonization of the annotation, however, is often problematic, especially when highly subjective annotation tasks are performed, e.g., involving pragmatics. Discarding and averaging discordant opinions carries the risk of losing the rich knowledge coming from different annotators' perspectives.
The Perspectivist Manifesto [1] represents an effort towards preserving the disagreement in data-centered AI, while strong perspectivism [2] extends this principle to the full Natural Language Processing pipeline.
In this talk, I will present the new perspectivist paradigm, focusing on NLP, and a few recent works exploring its implications on data annotation, model evaluation, and interoperability.
Much of contemporary Artificial Intelligence is based on supervised machine learning, a methodology that leverages large manually annotated datasets. The harmonization of the annotation, however, is often problematic, especially when highly subjective annotation tasks are performed, e.g., involving pragmatics. Discarding and averaging discordant opinions carries the risk of losing the rich knowledge coming from different annotators' perspectives.
The Perspectivist Manifesto [1] represents an effort towards preserving the disagreement in data-centered AI, while strong perspectivism [2] extends this principle to the full Natural Language Processing pipeline.
In this talk, I will present the new perspectivist paradigm, focusing on NLP, and a few recent works exploring its implications on data annotation, model evaluation, and interoperability.
[1] https://pdai.info
[2] http://www.itais.org/itais2021-proceedings/pdf/21.pdf
[2] http://www.itais.org/itais2021-proceedings/pdf/21.pdf
Short Bio:
Valerio Basile is an Assistant Professor at the Computer Science Department of the University of Turin, Italy, member of the Content-centered Computing group and the Hate Speech Monitoring lab. His work spans across several areas such as: formal representations of meaning, linguistic annotation, natural language generation, commonsense knowledge, semantic parsing, sentiment analysis, and hate speech detection, perspectives and bias in supervised machine learning, from data creation to system evaluation.
The session is open to everyone. Please register at