What is it about?

Imagine shopping online and seeing a "frequently bought together" package, or opening Spotify to find a perfectly curated music playlist. Instead of recommending just one single item, these platforms recommend a "bundle." This article serves as a comprehensive guide to understanding how these Bundle Recommendation Systems work. We reviewed years of research and categorized these systems into two main types: "discriminative" (systems that pick the best existing bundle for you) and "generative" (systems that create a brand-new, customized bundle from scratch).

Featured Image

Why is it important?

We live in a digital world overflowing with choices, which often leads to decision fatigue. Picking items one by one—like selecting individual clothing pieces for an outfit, or searching for separate flights and hotels for a trip—can be exhausting. Bundle recommendations solve this by offering cohesive, ready-to-use sets that save time and mental energy. For businesses, this not only makes customers much happier but also drives higher sales through natural cross-selling. As AI continues to evolve, understanding how to build smarter, more personalized bundles is becoming essential for e-commerce, entertainment, and travel industries.

Read the Original

This page is a summary of: A Survey on Bundle Recommendation: Methods, Applications, and Challenges, ACM Computing Surveys, March 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3802820.
You can read the full text:

Read

Contributors

The following have contributed to this page