In this paper we discussed conventional approaches to building sequential recommender systems and implemented BERT4Rec and LSTM-based models for two type of tasks in sequential recommendation: Relevance Prediction and Next Item Prediction. The usage of these models allowed to make use of sequential context, that is extremely important to simultaneously model the short-term interest of users in the session. Moreover, the usage of these models resulted in overall 0.5% increase in the quality of recommendations.
-
Notifications
You must be signed in to change notification settings - Fork 0
"Sequential recommender models" 3rd year coursework at Higher School of Economics
License
whiteRa2bit/coursework
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
"Sequential recommender models" 3rd year coursework at Higher School of Economics
Topics
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published