Chapter 09 · Sustainability & AISustainability & AI

Recommender system

Meaning statusEstablishedSource recordDirect document linkedWhy these are different

Definition

Software that ranks and suggests items — products, content, connections — to users based on predicted preference, typically learned from behaviour at scale.

References

SpringerRicci et al. (eds) — Recommender Systems Handbook, 2nd ed. (2022)

This source provides part of the technical or institutional basis for the definition.

IEEE ComputerKoren — The BellKor Solution to the Netflix Grand Prize (2009)

This source supports the explanation of how the term is applied, measured or governed in practice.

Overview

What it means

Recommenders combine techniques such as collaborative filtering and learned ranking, surveyed comprehensively in the Recommender Systems Handbook. They mediate a large share of online consumption, attention and purchasing.

How it is used

Relevance to sustainability runs through consumption patterns: recommenders steer what people buy, watch and read. They can be tuned to surface lower-impact products or credible information — or to maximise engagement regardless of consequence, including amplifying misleading environmental content.

Why it matters

Recommenders are among the most consequential algorithms for sustainability precisely because they are invisible infrastructure of demand. Their objective functions — what they are optimised to maximise — are a legitimate object of sustainability scrutiny.