Figure - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Efficiency analysis of item scoring methods. mRT is the Median Response Time, measured in milliseconds; SAS is the SASRec model and BERT is the gBERT4Rec model.
Source publication
Transformer-based recommender systems, such as BERT4Rec or SASRec, achieve state-of-the-art results in sequential recommendation. However, it is challenging to use these models in production environments with catalogues of millions of items: scaling Transformers beyond a few thousand items is problematic for several reasons, including high model me...
Context in source publication
Context 1
... of PQTopK and other scoring methods. Table 3 reports effectiveness and efficiency metrics for SASRec and BERT4Rec on both Booking.com and Gowalla datasets. We first observe that nDCG@10 values do not depend on the scoring method, as all algorithms compute the same score distribution. ...Similar publications
In modern social media, recommender systems (RecSys) rely on the click-through rate (CTR) as the standard metric to evaluate user engagement. CTR prediction is traditionally framed as a binary classification task to predict whether a user will interact with a given item. However, this approach overlooks the complexity of real-world social modeling,...
