liveBook · Manning

Personalized relevance

As explained in Hidden Influences

The estimated usefulness or interest of an item for a particular user.

When a social feed fills with updates from people you rarely interact with, personalized relevance helps bring likely-interesting posts forward. It estimates what may interest a particular user from observed behavior rather than an explicit list of preferences. section 1.2.2

Facebook’s early algorithmic feed

Introduced in 2007, EdgeRank prioritized newer posts and friends a user interacted with frequently. Seeing more of someone’s updates could prompt more interaction, creating a feedback loop. Unlike this straightforward weighting formula, current machine-learning recommendations are often harder to explain. section 1.2.2

How Hidden Influences covers it

How deep it goes
Explains EdgeRank’s weighting of recency and interaction, then contrasts it with less explainable modern recommendations.
Examples
Uses Facebook’s 2007 EdgeRank feed, which prioritized newer posts and frequently interacted-with friends.
What you can do after
Understand how observed interactions can shape feed rankings and reinforce further interaction.
How it connects
Connects personalized relevance to recommender systems, inferred preferences, and feedback loops.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Hidden Influences. The section links point to the parts of the book it draws on.