Recommenadation aided Caching using Combinatorial Multi-armed Bandits

2024-04-30

arXiv.org

AI Summary

In wireless networks with limited base-station cache capacity and uncertain user acceptance of recommended content, conventional decoupled caching and recommendation strategies suffer from suboptimal performance. Method: This paper proposes a joint caching and personalized recommendation optimization framework. It is the first to embed a recommendation mechanism into the Combinatorial Multi-Armed Bandit (CMAB) setting, designing a two-layer Upper Confidence Bound (UCB) algorithm that simultaneously learns content popularity and user-specific recommendation acceptance rates in an online manner. A unified cache–recommendation co-decision model is formulated to jointly optimize content placement and delivery policies. Contribution/Results: The algorithm provides a rigorous theoretical upper bound on cumulative regret. Numerical experiments demonstrate significant improvements in cache hit ratio over traditional disjoint baselines, validating the effectiveness, convergence, and practicality of the proposed joint modeling approach.

Technology Category

• Machine Learning: Online Learning & Bandits
• Search and Optimization: Learning to Search
• Data Mining & Knowledge Management: Recommender Systems

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Web Mining and Content Analysis: Content-based information diffusion
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

We study content caching with recommendations in a wireless network where the users are connected through a base station equipped with a finite-capacity cache. We assume a fixed set of contents with unknown user preferences and content popularities. The base station can cache a subset of the contents and can also recommend subsets of the contents to different users in order to encourage them to request the recommended contents. Recommendations, depending on their acceptability, can thus be used to increase cache hits. We first assume that the users' recommendation acceptabilities are known and formulate the cache hit optimization problem as a combinatorial multi-armed bandit (CMAB). We propose a UCB-based algorithm to decide which contents to cache and recommend and provide an upper bound on the regret of this algorithm. Subsequently, we consider a more general scenario where the users' recommendation acceptabilities are also unknown and propose another UCB-based algorithm that learns these as well. We numerically demonstrate the performance of our algorithms and compare these to state-of-the-art algorithms.

Problem

• Wireless Networks
• Content Recommendation
• Limited Storage

Innovation

• Multi-Armed Bandit Approach
• Content Caching and Recommendation
• Dynamic Strategy Adjustment