Editing
Recommendation Systems
(section)
Jump to navigation
Jump to search
Warning:
You are not logged in. Your IP address will be publicly visible if you make any edits. If you
log in
or
create an account
, your edits will be attributed to your username, along with other benefits.
Anti-spam check. Do
not
fill this in!
== <span style="color: #FFFFFF;">Remembering</span> == * '''Collaborative filtering''' β Recommends items based on the preferences of similar users or by finding items similar to those a user has liked. * '''Content-based filtering''' β Recommends items based on their attributes and the user's preference history (not other users). * '''Hybrid recommender''' β Combines collaborative and content-based approaches to leverage strengths of both. * '''User-item matrix''' β A matrix where rows are users, columns are items, and values are ratings or interaction signals; typically very sparse. * '''Matrix factorization''' β Decomposing the user-item matrix into lower-dimensional user and item embedding matrices (SVD, ALS). * '''Implicit feedback''' β Interaction signals without explicit ratings: clicks, views, time spent, purchases. Most real-world data is implicit. * '''Explicit feedback''' β Direct ratings (1-5 stars); less common in practice but higher signal. * '''Cold start problem''' β The difficulty of making recommendations for new users or new items with little or no interaction history. * '''Long tail''' β The vast number of niche items with few interactions; recommenders must balance popular items with relevant niche ones. * '''Click-through rate (CTR)''' β The fraction of recommended items that users click; a primary metric for recommendation quality. * '''Two-tower model''' β A neural architecture with separate encoders for users and items that are then combined (common in industrial recommenders). * '''Approximate Nearest Neighbor (ANN)''' β Fast similarity search in embedding space used to retrieve candidate items at scale (FAISS, ScaNN). * '''Explore-exploit tradeoff''' β Balancing recommending items predicted to be liked (exploit) vs. items with uncertain engagement potential (explore). * '''Diversity''' β Ensuring recommended items are not all from the same narrow category; improves user satisfaction and avoids filter bubbles. </div> <div style="background-color: #006400; color: #FFFFFF; padding: 20px; border-radius: 8px; margin-bottom: 15px;">
Summary:
Please note that all contributions to BloomWiki may be edited, altered, or removed by other contributors. If you do not want your writing to be edited mercilessly, then do not submit it here.
You are also promising us that you wrote this yourself, or copied it from a public domain or similar free resource (see
BloomWiki:Copyrights
for details).
Do not submit copyrighted work without permission!
Cancel
Editing help
(opens in new window)
Navigation menu
Personal tools
Not logged in
Talk
Contributions
Create account
Log in
Namespaces
Page
Discussion
English
Views
Read
Edit
View history
More
Search
Navigation
Main page
Recent changes
Random page
Help about MediaWiki
Tools
What links here
Related changes
Special pages
Page information