This page lists my current research seminar and other standing talks. I can give any of these lectures with relatively short notice. I list them below with approximate time slot lengths, including questions; I can adjust to fit your schedule.
I will also consider invitations to give talks on the social impact of recommender systems, particularly with regards to bias and discrimination; principles of recommender systems; and evaluating recommender systems.
This is my normal research seminar, primarily targeted for researchers and advanced undergraduate students in computer science, information science, and related disciplines. I am continuously updating this talk; its last major rewrite was Fall 2018.
Title: User, Agent, Subject, Spy: Information Systems for Human Flourishing
Length: 50 minutes (plus questions)
Every day, information access systems mediate our experience of the world beyond our immediate senses. Google helps us find what we seek, Amazon and Netflix recommend things for us to buy and watch, Apple News gives us the day's events, and BuzzFeed guides us to related articles. These systems deliver immense value, but also have profound influence over how we experience information and the resources and perspectives we see.
I will report on several projects using data analysis and simulation to illuminate the ways in which the algorithms underlying these systems, such as collaborative filters, respond to various human factors reflected in the input and response data that drive them. We probe how recommender system experiments respond to systematically missing data that bias the experimental results, how they interact with patterns in their input that may reflect historical or ongoing discrimination, and other individual and social human concerns.
I will also discuss system-building and experimental work we are engaged in to understand how users perceive and interact with the recommendation and retrieval systems they use, and to develop novel systems that empower users to leverage information to improve their lives and work. Our current project in this space is tool that helps K–12 teachers find new readings to enhance their classroom instruction.
All these projects are a part of our broad vision to ensure information systems are responsive to the needs and well-being of the people they affect.
This talk presents a higher-level view of my research work, targeted at undergraduate students who are not necessarily in computer science.
Title: Making Information Systems Good for People
Length: 50 minutes (plus questions)
Every day, we interact with countless information systems enhanced by artificial intelligence: Google helps us find what we seek, Amazon and Netflix recommend things for us to buy and watch, Apple News gives us the day's events, BuzzFeed guides us to related articles, and VISA assesses whether our morning coffee is a fraudulent purchase. These systems deliver immense value, but also have profound influence over how we experience information and the resources and perspectives we see.
In this talk, I will explain the basic principles of modern data-driven artificial intelligence systems, and present some of the work that I and others are doing to measure their human impact and ensure that they are good for the people they affect: that they provide value, minimize harm, and produce outcomes that treat people fairly. I will pay particular attention to recommender systems, the algorithms that suggest products and articles, and how they do or do not meet their users' needs.
This is a shorter, higher-level talk on ethics and user engagement in designing and evaluating recommender systems.
Title: Responsible Recommendation
Length: 20 minutes
I gave this talk at the Brussels Data Ethics meeting.
Dr. Michael Ekstrand is an assistant professor in the Department of Computer Science at Boise State University, where he co-directs the People and Information Research Team (PIReT, pronounced ‘pirate’). He studies human-computer interaction and artificial intelligence, working to ensure information access systems — particularly recommender systems — are good for the people they affect. He received his Ph.D in 2014 from the University of Minnesota, building tools reproducible research and examining user-relevant differences in recommender algorithms with the GroupLens research group; leads the LensKit open-source software project for enabling high-velocity reproducible research in recommender systems; and co-created the Recommender Systems specialization on Coursera with Joseph A. Konstan from the University of Minnesota. In 2018 he receieved the NSF CAREER award to fund his study of recommender system response to biases in input data and experimental protocols and their projected impact under various technical and sociological conditions.