AI Spotlight #1, Ingeniería de Inteligencia Artificial, Universidad Peruana de Ciencias Aplicadas (UPC)
Online (Microsoft Teams)
Inaugural talk of AI Spotlight, a new speaker series from UPC’s AI Engineering programme, on using language models to adapt text to different readers.
SEPLN Online Talks
Online (Zoom)
Invited talk on automatic text simplification adapted to different reader proficiency levels, covering readability control, complexity representation, and evaluation challenges.
Cognitive Science Seminar Series
Queen Mary University of London (online)
I present the first meta-evaluation of automatic metrics for Automatic Sentence Simplification focused on simplicity.
Research Seminar in Natural Language Processing
University of Wolverhampton (online)
I provide an overview of how automatically-simplified texts are currently evaluated, highlighting issues with resources and metrics for automatic evaluation, and proposing solutions based on recent work.
Simplify Language – Capture Audience
ELRC Technical Workshop (online)
I discuss how evaluating automatic simplifications is difficult, since we do not have metrics that measure simplicity with high confidence, nor human judgements majorly agree in what simple is.
BeyondMinds - AI Talks
BeyondMinds (online)
I present work that attempts to incorporate the multi-operation nature of Text Simplication into automatic models and evaluation resources.
SheffieldNLP Seminar
University of Sheffield (online)
I present the results of the first meta-evaluation of automatic metrics for Sentence Simplification, focused on simplicity judgements.
NLP with Friends
NLP with Friends (online)
I explain the limitations of evaluation resources for Text Simplification when assessing multi-operation simplifications, and present some of our latest work on overcoming them.
5th Annual Amazon Graduate Research Symposium
Amazon (Seattle)
I review our results on a manual analysis of professionally-produced simplifications, which sheds light on the transformation operations that humans perform at the document-level.