Solving the Content Gap in Roblox Game Recommendations: LLM-Based Profile Generation and Reranking

• DOI: 10.48550/arxiv.2502.06802
• Publication Date: 2025-02-01
• Authors: Chen Wang; Xiaokai Wei; Yi Jiang; Frank Ong; Kai Gao; Xiao‐Ying Yu; Hui Zheng; Se-eun Yoon; Philip S. Yu; Minghui Gong
• Keywords: Content (measure theory)

With the vast and dynamic user-generated content on Roblox, creating effective game recommendations requires a deep understanding of content. Traditional recommendation models struggle with inconsistent sparse nature text features such as titles descriptions. Recent advancements in large language (LLMs) offer opportunities to enhance systems by analyzing in-game data. This paper addresses two challenges: generating high-quality, structured for games without extensive human annotation, validating these ensure they improve relevance. We propose an approach that extracts uses LLMs infer attributes genre gameplay objectives from raw player interactions. Additionally, we introduce LLM-based re-ranking mechanism assess effectiveness generated features, enhancing personalization user satisfaction. Beyond recommendations, our supports applications engagement-based integrity detection, already deployed production. scalable framework demonstrates potential quality Roblox adapt its unique, ecosystem.