Year: 2025
Type: article
Abstract: Linear shallow autoencoders have gained popularity in collaborative filtering benchmarks for recommender systems due to their simplicity and strong performance. However, models like EASE struggle to scale to datasets with a large number of items. This paper addresses this limitation by evaluating scalable variants of shallow linear autoencoders, namely, the recently proposed ELSA and SANSA, on large-scale datasets representative of real-world applications. We further propose improvements to enhance their scalability on large, sparse datasets. Our evaluation goes beyond standard offline and online performance metrics, incorporating key industrial considerations such as training time and memory usage. By analyzing the effects of data sparsity and catalog size, we provide practical guidance to select the most suitable model for a given dataset. We release the source code, training logs, and detailed instructions to reproduce our experiments at https://github.com/recombee/SparseELSA.
DOI: 10.1145/3748335
Authors: Vojtěch Vančura, Petr Kasalický, Rodrigo Alves, Pavel Kordík
Institution: Czech Technical University in Prague
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4412363075