This is a project done to study the effect of of popularity bias and its effect coming to the conclusion on how to remove it. With help of lighfm i used Movies-lens dataset to create a simple recommender system and recommended 20 movies to a random user. It was found that all the recommended movies was only the top rated ones but with good genere similarity. By comparing the embedding of the recommended to the longtail ones i recommended 5 movies which have low interaction data. So at last we have 20 movies in which 5 is long tail. So in this way i eliminated the popularity bias.
-
Notifications
You must be signed in to change notification settings - Fork 0
Ed-here/Recommender-System
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
No description, website, or topics provided.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published