Integrating Predictive Models into Two-Sided Recommendations: A Matching-Theoretic Approach

2026-02-23

AI Summary

This study addresses the limitations of conventional two-sided recommendation systems, where reliance solely on matching probabilities often leads to overexposure of popular users, system congestion, and inflated matching efficiency. To mitigate these issues, the authors formulate recommendation as a many-to-many matching problem and introduce an “effective date” metric that integrates login activity, likes, and mutual selection probabilities. They propose the Exposure-Constrained Deferred Acceptance (ECDA) algorithm, which optimizes match quality under receiver-side capacity constraints by limiting exposure based on expected interaction volume rather than user count. Evaluated on real-world data from a Japanese dating platform, ECDA significantly increases both the number of effective dates and the probability of receivers securing dates, while preserving downstream user engagement and ensuring fairness.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Data Mining & Knowledge Management: Recommender Systems
• Multiagent Systems: Mechanism Design

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Economics, Online Markets and Human Computation: Economics and fairness of platforms and recommendation systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Two-sided platforms must recommend users to users, where matches (termed \emph{dates} in this paper) require mutual interest and activity on both sides. Naive ranking by predicted dating probabilities concentrates exposure on a small subset of highly responsive users, generating congestion and overstating efficiency. We model recommendation as a many-to-many matching problem and design integrators that map predicted login, like, and reciprocation probabilities into recommendations under attention constraints. We introduce \emph{effective dates}, a congestion-adjusted metric that discounts matches involving overloaded receivers. We then propose \emph{exposure-constrained deferred acceptance} (ECDA), which limits receiver exposure in terms of expected likes or dates rather than headcount. Using production-grade predictions from a large Japanese dating platform, we show in calibrated simulations that ECDA increases effective dates and receiver-side dating probability despite reducing total dates. A large-scale regional field experiment confirms these effects in practice, indicating that exposure control improves equity and early-stage matching efficiency without harming downstream engagement.

Problem

• two-sided recommendations
• congestion
• exposure control
• matching efficiency
• equity

Innovation

• two-sided recommendations
• matching theory
• exposure-constrained deferred acceptance
• effective dates
• congestion-aware matching