Specializations
- Responsible and Human-Centered AI
- Computational Social Science
- Natural Language Processing (NLP)
Research Areas
Biography
Belén Saldías is an Assistant Teaching Professor at the University of Washington Information School.
Her research focuses on the responsible, human-centered design and evaluation of machine learning systems, emphasizing human-AI collaboration. In her Ph.D. at MIT, she explored how language models can support decentralized governance and increase user agency online. Saldías is committed to making the internet safer and more empowering, especially for children and marginalized communities, integrating machine learning and natural language processing with social science insights.
As an educator, Saldías teaches courses addressing pressing social issues such as Generative AI Ethics and Data Justice, alongside technical topics including Databases and Information Visualization. She views teaching as a relational practice, emphasizing self-awareness, strategic learning, rigor, and practical adaptability.
Saldías earned her Ph.D. at the MIT Media Lab and the Center for Constructive Communication. Previously, she held positions at Harvard University, Google Research, Pontificia Universidad Católica de Chile, and several AI-focused companies and startups.
Education
- Ph D, Media Arts & Sciences, Massachusetts Institute of Technology
- MS, Machine Learning, Pontificia Universidad Católica de Chile
- BS, Industrial Engineering Information Technology, Pontificia Universidad Católica de Chile
Awards
- MIT Women of Excellence Award - MIT, 2023
- Award for Highest Teaching Quality - Pontificia Universidad Católica de Chile, 2018
- Award for Best Computer Science Thesis - Pontificia Universidad Católica de Chile, 2017
- Award for Exceptional Quality as Teaching Assistant - Pontificia Universidad Católica de Chile, 2015
Publications and Contributions
-
Conference PosterDesigning Child-Centered Content Exposure and Moderation (2024)10th International Conference on Computational Social Science (IC2S2)
-
Conference PaperSenseMate: An Accessible and Beginner-Friendly Human-AI Platform for Qualitative Data Analysis (2024)Proceedings of the 29th International Conference on Intelligent User Interfaces (IUI), pp. 922-939
-
Conference PaperToward More Effective Human Evaluation for Machine Translation (2022)Proceedings of the 2022 ACL Workshop on Human Evaluation of NLP Systems (HumEval), pp. 76–89
-
Conference PosterExploring aspects of similarity between spoken personal narratives by disentangling them into narrative clause types (2020)WiML at ICML
-
Conference PaperExploring aspects of similarity between spoken personal narratives by disentangling them into narrative clause types (2020)Proceedings of the First Joint ACL Workshop on Narrative Understanding, Storylines, and Events (NUSE 2020), pp. 78-86
-
Conference PosterUnderstanding story similarity for personal narratives (2020)14th Women in Machine Learning (WiML) Workshop co-located with NeurIPS
-
Conference PaperA Full Probabilistic Model for Yes/No Type Crowdsourcing in Multi-Class Classification (2019)Proceedings of the 2019 SIAM International Conference on Data Mining (SDM), pp. 756-764
-
Conference PaperGobo: A System for Exploring User Control of Invisible Algorithms in Social Media (2019)ACM Conference on Computer Supported Cooperative Work and Social Computing (CSCW 2019), pp. 151-155
-
Conference PaperTweet Moodifier: Towards Giving Emotional Awareness to Twitter Users (2019)Proceedings of the 8th International Conference on Affective Computing and Intelligent Interaction (ACII 2019)
Presentations
-
Machine Learning Research and Technology
(2024)
Korea Advanced Institute of Science and Technology - South Korea
-
Intro to Machine Learning and Generative AI
(2023)
MIT MISTI and MIT GSL - Cambridge, MA
-
Decentralized content moderation
(2023)
MIT - Cambridge, MA
-
Does AI Hold the Same Values We Do?
(2023)
TEDx Bentley U - Waltham, MA
-
From Words to Online Content Moderation
(2023)
MIT – Chile Research Workshop - Cambridge, MA
-
LLMs and shaping healthier online spaces
(2023)
Generative AI for Constructive Communication, MIT - Cambridge, MA
-
LLMs and Us: How well can LLMs represent my values?
(2023)
Ars Electronica, Johannes Kepler University Linz, Austria - Linz, Austria
-
LLMs, responsible sentiment analysis and content moderation
(2023)
MIT - Cambridge, MA
-
Intro to Machine Learning
(2022)
MIT MISTI and MIT GSL - Cambridge, MA
-
Machine Learning in Industry
(2022)
Agrosuper - Chile
-
Probabilistic Models for Clustering
(2022)
Pontificia Universidad Católica de Chile (PUC) - Santiago, Chile
-
Responsible affect manipulation and social media
(2022)
MIT - Cambridge, MA
-
Working with Teaching Assistants
(2022)
MIT Teaching + Learning Lab - Cambridge, MA
-
Developing a Research Thesis
(2021)
Introduction to Media Arts and Sciences at MIT - Cambridge, MA
-
Ethical online experiments for mental health
(2021)
MIT - Cambridge, MA
-
Advances in Human-Computer Interaction
(2020)
Social Computing & Visualization Group Pontificia Universidad Católica de Chile (PUC) - Santiago, Chile
-
Data Science in Industry
(2020)
Universidad of Concepción, Chile - Concepción, Chile
-
Innovation and Entrepreneurship
(2020)
Universidad de La Sabana, Colombia - Chía, Colombia
-
Mitigating Bias in Analytics
(2020)
Women in Analytics and Tableau - Virtual
-
Research in Machine Learning
(2020)
Pontificia Universidad Católica de Chile (PUC) CS/DCC PhD Seminar - Santiago, Chile
-
Towards Children-Aware Machine Learning with a Focus on NLP Challenges and Applications
(2020)
1st WiML Un-Workshop co-located with ICML - Virtual
-
Advances in Human-Computer Interaction
(2019)
Social Computing & Visualization Group Pontificia Universidad Católica de Chile (PUC) - Santiago, Chile
-
Ethical online experiments with human subjects
(2019)
MIT - Cambridge, MA
-
End-to-end machine learning in industry
(2018)
Women in Data Science - Santiago, Chile