Collaborative filtering
Collaborative filtering is a recommendation technique that infers preferences or associations from interaction patterns among users, items, or other entities.
AI-Powered Search
Collaborative recommendation predicts items or content a user may prefer from the behavior and preferences of similar users.
- How it is used
- Personalizes recommendations and search results using patterns in user–item interactions.
- How deep it goes
- Develops the workflow from interaction signals and user–item matrices through sparse-data factorization, latent features, and recommendation generation.
- Examples
- Illustrates overlapping behavior with fertilizer shoppers who also interact with soil or mulch, rather than an unrelated tool cluster.
- What you can do after
- Build it from scratch using user–item interaction signals to learn preferences and generate item recommendations.
Read how AI-Powered Search explains collaborative filtering →
The Art of AI Product Development
A collaborative-filtering approach that compares users and recommends items based on the preferences of similar users.
- How it is used
- Deploys existing behavior data to personalize product or content recommendations and support an initial personalization feature.
- How deep it goes
- Defines user-based and item-based approaches, then extends treatment to testing, feedback, metrics, and cold-start limitations.
- Examples
- Uses shared shoe purchases, product interactions, click-through rates, conversions, browsing time, and thumbs-up or thumbs-down feedback.
- What you can do after
- Apply collaborative filtering to recommend items, compare filtering strategies, and refine results through metrics, experiments, and feedback.
Read how The Art of AI Product Development explains collaborative filtering →
Knowledge Graphs and LLMs in Action
Collaborative filtering is a recommendation technique that infers preferences or associations from patterns across users, items, or entities.
- How it is used
- Infers drug–disease, drug–target, and treatment associations for pharmaceutical analysis and drug repurposing.
- How deep it goes
- Defines applied uses and coverage limitations, but stops short of algorithmic implementation details.
- Examples
- Uses drug–disease associations, drug–target interactions, and drug–disease treatments as prediction targets.
- What you can do after
- Evaluate association predictions while recognizing that graph coverage constrains the drugs and diseases it can address.
Read how Knowledge Graphs and LLMs in Action explains collaborative filtering →
First-Party Data Activation
A recommendation technique that infers a user's interests from patterns in the preferences or behavior of multiple users.
- How it is used
- Positions collaborative filtering as a privacy-enhancing approach for personalizing content and interactions.
- How deep it goes
- Defines its privacy role but stops short of explaining algorithms, data structures, or implementation procedures.
- What you can do after
- Apply it via privacy-enhancing personalization while limiting stored personal data and intrusive tracking.
Read how First-Party Data Activation explains collaborative filtering →
Related concepts
- recommender systemkind of · A recommender system is software that predicts, selects, or ranks items, content, or actions for users based on available information about users, items, or context.
- content-based recommendercontrasts with · A content-based recommender suggests items whose features resemble those of items matching a user’s interests or prior interactions.
- behavioral datauses · Behavioral data is information recording the observable actions, interactions, responses, or usage patterns of people interacting with a product, service, or system.
- matrix factorizationuses · Matrix factorization is the decomposition or approximation of a matrix as a product of simpler, often lower-dimensional or structured matrices.
- data-driven approachkind of · A data-driven approach bases decisions, conclusions, or system behavior substantially on collected or observed data rather than primarily on intuition or assumptions.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.