Scaling Session-Based Transformer Recommendations using Optimized Negative Sampling and Loss Functions

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wilm, Timo, Normann, Philipp, Baumeister, Sophie, Kobow, Paul-Vincent
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915218010931200
author Wilm, Timo
Normann, Philipp
Baumeister, Sophie
Kobow, Paul-Vincent
author_facet Wilm, Timo
Normann, Philipp
Baumeister, Sophie
Kobow, Paul-Vincent
contents This work introduces TRON, a scalable session-based Transformer Recommender using Optimized Negative-sampling. Motivated by the scalability and performance limitations of prevailing models such as SASRec and GRU4Rec+, TRON integrates top-k negative sampling and listwise loss functions to enhance its recommendation accuracy. Evaluations on relevant large-scale e-commerce datasets show that TRON improves upon the recommendation quality of current methods while maintaining training speeds similar to SASRec. A live A/B test yielded an 18.14% increase in click-through rate over SASRec, highlighting the potential of TRON in practical settings. For further research, we provide access to our source code at https://github.com/otto-de/TRON and an anonymized dataset at https://github.com/otto-de/recsys-dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2307_14906
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Scaling Session-Based Transformer Recommendations using Optimized Negative Sampling and Loss Functions
Wilm, Timo
Normann, Philipp
Baumeister, Sophie
Kobow, Paul-Vincent
Information Retrieval
Artificial Intelligence
Machine Learning
This work introduces TRON, a scalable session-based Transformer Recommender using Optimized Negative-sampling. Motivated by the scalability and performance limitations of prevailing models such as SASRec and GRU4Rec+, TRON integrates top-k negative sampling and listwise loss functions to enhance its recommendation accuracy. Evaluations on relevant large-scale e-commerce datasets show that TRON improves upon the recommendation quality of current methods while maintaining training speeds similar to SASRec. A live A/B test yielded an 18.14% increase in click-through rate over SASRec, highlighting the potential of TRON in practical settings. For further research, we provide access to our source code at https://github.com/otto-de/TRON and an anonymized dataset at https://github.com/otto-de/recsys-dataset.
title Scaling Session-Based Transformer Recommendations using Optimized Negative Sampling and Loss Functions
topic Information Retrieval
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2307.14906