Skip to main content

Types of Recommender Systems

Group Recommender System​

Recommend items to a group of users, e.g., group dining

Multi-Stakeholder Recommender System​

Produce recommendations by considering multiple stakeholders, e.g., buyers and sellers on eBay

Multi-Task Recommender System​

Build joint learning model by considering multiple tasks, e.g., RecSys + opinion texts

Content-based Recommender System​

Below figure shows the high-level architecture of a CBRS, one of many possible architectures.

Untitled

Collaborative Filtering Recommender System​

A graph-powered collaborative filtering recommender system

Untitled

Session-based Recommender System​

Untitled

Context-aware Recommender System​

Incorporate context info (time, location, etc) into RecSys

Untitled

Based on how the contextual information, the current user, and the current item are used during the recommendation process, context-aware recommendation systems can take one of the three forms shown below:

Untitled

Hybrid Recommender System​

Untitled

Hybridization design techniques

Untitled