Unpersonalized Recommenders¶
The Unpersonalized Recommenders module of WarpRec serves as a collection of simple baselines. These models do not learn from individual user interactions but instead rely on global statistics or random selection. They are essential for producing reference metric scores against which the performance of actual (personalized) models can be compared.
In the following sections, you will find the list of available unpersonalized models within WarpRec.
API Reference
For class signatures, parameters, and source code, see the Unpersonalized API Reference.
Pop¶
Pop: Recommends the most popular items overall. This model helps assess whether other recommenders are biased towards popularity.
Random¶
Random: Recommends items at random. This model defines a lower bound for performance metrics, serving as a sanity check during evaluation.
Cross-Framework Evaluation
Looking to evaluate recommendation lists from external frameworks? See Cross-Framework Evaluation.