๐ค AI Summary
To address memory overflow and poor scalability caused by fixed-size embeddings in recommender systems, this paper proposes a budget-controllable, table-level embedding optimization method. Unlike existing instance-wise search approaches or manually tuned lightweight embedding schemes, our method introduces a novel table-level action modeling and set representation learning framework, incorporating an action suitability predictor to jointly optimize embedding dimensions across all feature tablesโglobally, efficiently, and strictly under user-specified memory constraints. By integrating set neural networks with memory-aware architecture search, the approach eliminates the need for heuristic trade-off coefficients. Evaluated on two real-world datasets with three mainstream recommendation models, our method achieves state-of-the-art performance, significantly improving both recommendation accuracy and memory efficiency while rigorously satisfying multiple predefined memory budgets.
๐ Abstract
At the heart of contemporary recommender systems (RSs) are latent factor models that provide quality recommendation experience to users. These models use embedding vectors, which are typically of a uniform and fixed size, to represent users and items. As the number of users and items continues to grow, this design becomes inefficient and hard to scale. Recent lightweight embedding methods have enabled different users and items to have diverse embedding sizes, but are commonly subject to two major drawbacks. Firstly, they limit the embedding size search to optimizing a heuristic balancing the recommendation quality and the memory complexity, where the trade-off coefficient needs to be manually tuned for every memory budget requested. The implicitly enforced memory complexity term can even fail to cap the parameter usage, making the resultant embedding table fail to meet the memory budget strictly. Secondly, most solutions, especially reinforcement learning based ones derive and optimize the embedding size for each each user/item on an instance-by-instance basis, which impedes the search efficiency. In this paper, we propose Budgeted Embedding Table (BET), a novel method that generates table-level actions (i.e., embedding sizes for all users and items) that is guaranteed to meet pre-specified memory budgets. Furthermore, by leveraging a set-based action formulation and engaging set representation learning, we present an innovative action search strategy powered by an action fitness predictor that efficiently evaluates each table-level action. Experiments have shown state-of-the-art performance on two real-world datasets when BET is paired with three popular recommender models under different memory budgets.
