Optimal Dataset Size for Recommender Systems: Evaluating Algorithms' Performance via Downsampling

TL;DR

This study evaluates how dataset downsampling affects the performance and energy efficiency of recommender systems, demonstrating significant reductions in runtime and carbon emissions while maintaining competitive recommendation quality.

Contribution

It introduces a systematic analysis of dataset downsampling as a method to optimize energy efficiency in recommender systems, with empirical evidence across multiple datasets and algorithms.

Findings

30% downsampling reduces runtime by 52% and carbon emissions by up to 51.02 KgCO2e.

Algorithms retain 81% of full dataset performance at 50% training data.

Some configurations outperform full datasets in recommendation quality.

Abstract

This thesis investigates dataset downsampling as a strategy to optimize energy efficiency in recommender systems while maintaining competitive performance. With increasing dataset sizes posing computational and environmental challenges, this study explores the trade-offs between energy efficiency and recommendation quality in Green Recommender Systems, which aim to reduce environmental impact. By applying two downsampling approaches to seven datasets, 12 algorithms, and two levels of core pruning, the research demonstrates significant reductions in runtime and carbon emissions. For example, a 30% downsampling portion can reduce runtime by 52% compared to the full dataset, leading to a carbon emission reduction of up to 51.02 KgCO2e during the training of a single algorithm on a single dataset. The analysis reveals that algorithm performance under different downsampling portions depends on factors like dataset characteristics, algorithm complexity, and the specific downsampling configuration (scenario dependent). Some algorithms, which showed lower nDCG@10 scores compared to higher-performing ones, exhibited lower sensitivity to the amount of training data, offering greater potential for efficiency in lower downsampling portions. On average, these algorithms retained 81% of full-size performance using only 50% of the training set. In certain downsampling configurations, where more users were progressively included while keeping the test set size fixed, they even showed higher nDCG@10 scores than when using the full dataset. These findings highlight the feasibility of balancing sustainability and effectiveness, providing insights for designing energy-efficient recommender systems and promoting sustainable AI practices.