Efficient Sequential Recommendation for Long Term User Interest Via Personalization

TL;DR

This paper introduces a personalized sequential recommendation method that compresses user histories into learnable tokens, significantly reducing computational costs while maintaining high accuracy, applicable to existing transformer-based models.

Contribution

A novel approach that enhances efficiency in sequential recommendation by compressing user histories into learnable tokens, compatible with current transformer models.

Findings

Reduces computational costs in sequential recommendation models.

Maintains high recommendation accuracy with the proposed method.

Demonstrates effectiveness across multiple models and datasets.

Abstract

Recent years have witnessed success of sequential modeling, generative recommender, and large language model for recommendation. Though the scaling law has been validated for sequential models, it showed inefficiency in computational capacity when considering real-world applications like recommendation, due to the non-linear(quadratic) increasing nature of the transformer model. To improve the efficiency of the sequential model, we introduced a novel approach to sequential recommendation that leverages personalization techniques to enhance efficiency and performance. Our method compresses long user interaction histories into learnable tokens, which are then combined with recent interactions to generate recommendations. This approach significantly reduces computational costs while maintaining high recommendation accuracy. Our method could be applied to existing transformer based recommendation models, e.g., HSTU and HLLM. Extensive experiments on multiple sequential models demonstrate its versatility and effectiveness. Source code is available at \href{https://github.com/facebookresearch/PerSRec}{https://github.com/facebookresearch/PerSRec}.