MARCO：用于校准广告转化预测的点击意图分解
MARCO: Click-Intent Decomposition for Calibrated Ads Conversion Prediction
- Meta AI
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
MARCO框架通过按意图分解点击解决广告转化预测的校准偏差，经离线在线验证，部署后各意图校准近100%，每点击转化率提升2.80%，核心指标累计提升0.98%。
AI中文摘要：
并非所有点击都具有相同价值。工业广告排名将转化概率解耦为点击率（CTR）和点击后转化率（CVR），但将每一次点击视为相同事件。实际上，用户通过物理UI交互提供了免费的、自发的意图信号，同一广告上不同点击类型的实际转化率存在4倍差异。由于混淆了这些信号，标准CVR模型对高意图点击的预测偏低，对低意图点击的预测偏高，而这种偏差被近乎完美的整体校准所掩盖。我们提出MARCO（多意图广告排名组合优化，Multi-intent Ads Ranking Composition Optimization）框架，通过按意图分解每一次点击来解决这一偏差。利用记录的点击类型作为免费的行为标签，MARCO在同质群体上训练各意图的CVR头，并在服务阶段根据预测的意图分布组合各意图的CVR估计值。理论上，我们证明分解不会增加总体风险，给出了平方损失下的确切余量和部署损失下的非负性，并通过路由效率参数展示了其中多少余量会在服务阶段体现。由于总体最优得分保持不变，任何增益都是有限容量的估计和校准效应，我们通过离线和在线实验验证了这一点。为实现大规模部署，我们进一步将多曝光、多点击归因建模为具有类似强化学习（RL）回报估计的偏差-方差权衡的信用分配，表明最后点击归因、首次点击归因是生产约束下低偏差、低方差、确定性的选择，并推导了大规模端到端执行的三个一致性条件。在二元意图粒度下部署MARCO后，各意图校准被修正至约100%，每点击转化率提升2.80%，核心指标累计提升0.98%。
英文摘要：
Not all clicks are equal. Industrial ads ranking decouples conversion probability into click-through rate (CTR) and post-click conversion rate (CVR), yet treats every click as the same event. In reality, users provide a free, self-generated signal of intent through their physical UI interactions. Different click types on the same ad exhibit a 4-fold difference in actual conversion rates. By conflating these signals, the standard CVR model under-predicts high-intent clicks and over-predicts low-intent ones, which is a bias masked by near-perfect aggregate calibration. We propose MARCO (Multi-intent Ads Ranking Composition Optimization), a framework that resolves this bias by decomposing each click by intent. Using the logged click type as a free behavioral label, MARCO trains per-intent CVR heads on homogeneous populations, and at serving time composes their per-intent CVR estimates under a predicted distribution over intents. Theoretically, we prove that decomposition never raises population risk, give the exact headroom under squared loss and non-negativity under the deployed loss, and show through a routing-efficiency dial how much of it reaches serving. Because the population-optimal score is unchanged, any gain is a finite-capacity estimation and calibration effect that we validated both offline and online. For deployment at scale, we further cast multi-impression, multi-click attribution as credit assignment with a bias-variance tradeoff analogous to RL return estimation, showing last-impression, first-click attribution is the low-bias, low-variance, deterministic choice under production constraints, and derive three consistency conditions enforced end-to-end at scale. Deployed at binary intent granularity, MARCO corrects per-intent calibration to approximately 100%, lifts conversions per click by +2.80%, and drives +0.98% cumulative improvement in topline metrics.
