Recommender System
Chapter 7 · page 107
Definition
A recommender system is an algorithmic system designed to automatically offer content, products or services to a user based on their preferences, their past behaviour or that of users judged to be similar. It analyzes signals such as clicks, searches, purchases, views or time spent on certain content in order to anticipate what might hold the user’s attention. These systems are everywhere: on streaming platforms, online retail, social media and music apps.
Example
On a video streaming platform, the recommender system suggests films or series based on your viewing history and on the behaviour of users whose profiles are close to yours. On an e-commerce platform, it offers products related to your purchases or to those of similar customers.
Why it matters
Recommender systems filter the immense quantity of information available online and influence our choices, often without our being fully aware of it. Well calibrated, they help us discover content that is relevant or suited to our tastes. But because they also seek to maximize attention and engagement, they can favour emotional, polarizing or repetitive content. At scale, they therefore play a major role in how we inform ourselves, how we consume and even how we perceive the digital world.
How to cite this text
Vial, S. (2026). Recommender System. In A Living Lexicon of Artificial Intelligence (p. 107). Stéphane Vial, publisher, Montréal.
