Technology Category
Game Theory and Economic Paradigms: Adversarial LearningHumans and AI: Game Design — Procedural Content Generation & StorytellingMachine Learning: Online Learning & Bandits
Large-scale User Game Lifecycle Representation Learning
🤖 AI Summary
Modeling user interests in game advertising and recommendation is challenging due to game sparsity and popularity bias. Method: This paper proposes a User Game Lifecycle (UGL) representation learning framework, which innovatively incorporates the lifecycle concept, employs an inverse-probability masking strategy to mitigate popularity bias, and explicitly models multi-stage interest evolution via behavior reconstruction and separation of short-term and long-term interests. Technically, it integrates representation learning, sequential modeling, data augmentation, and bias correction. Contribution/Results: Offline experiments show AUC improvements of 1.83% on ad recommendation and 0.50% on in-game item recommendation. Online A/B testing demonstrates a 21.67% lift in CVR and a 0.82% increase in ARPU, significantly enhancing personalization efficacy in large-scale gaming scenarios.
Technology Category
Application Category
📝 Abstract
The rapid expansion of video game production necessitates the development of effective advertising and recommendation systems for online game platforms. Recommending and advertising games to users hinges on capturing their interest in games. However, existing representation learning methods crafted for handling billions of items in recommendation systems are unsuitable for game advertising and recommendation. This is primarily due to game sparsity, where the mere hundreds of games fall short for large-scale user representation learning, and game imbalance, where user behaviors are overwhelmingly dominated by a handful of popular games. To address the sparsity issue, we introduce the User Game Lifecycle (UGL), designed to enrich user behaviors in games. Additionally, we propose two innovative strategies aimed at manipulating user behaviors to more effectively extract both short and long-term interests. To tackle the game imbalance challenge, we present an Inverse Probability Masking strategy for UGL representation learning. The offline and online experimental results demonstrate that the UGL representations significantly enhance model by achieving a 1.83% AUC offline increase on average and a 21.67% CVR online increase on average for game advertising and a 0.5% AUC offline increase and a 0.82% ARPU online increase for in-game item recommendation.
Problem
Research questions and friction points this paper is trying to address.
Addressing game sparsity in user representation learning
Mitigating game imbalance from dominant popular titles
Enhancing game advertising and recommendation system performance
Innovation
Methods, ideas, or system contributions that make the work stand out.
Introducing User Game Lifecycle to enrich user behaviors
Proposing strategies to extract short and long-term interests
Using Inverse Probability Masking to address game imbalance
🔎 Similar Papers
No similar papers found.
