AdSight: Scalable and Accurate Quantification of User Attention in Multi-Slot Sponsored Search

2025-04-30

AI Summary

To address the challenge of accurately quantifying user attention on multi-slot search engine results pages (SERPs), this paper proposes the first Transformer-based sequence-to-sequence attention prediction framework specifically designed for multi-slot SERPs. The method decouples mouse trajectory modeling from slot-level semantic awareness, jointly embedding trajectory dynamics and slot-specific features, and simultaneously optimizes two complementary tasks: regression of gaze duration/fixation count and binary classification of slot attention. This dual-objective design significantly enhances cross-layout generalization. Evaluated on a large-scale real-world SERP dataset, the model reduces mean absolute error (MAE) in gaze time prediction by 32% and improves F1-score for slot attention detection by 19%. To our knowledge, this is the first work achieving high-accuracy, scalable attention modeling across multiple SERP slots, providing an interpretable and production-ready foundation for ad slot optimization and revenue strategy.

Technology Category

• Search and Optimization: Learning to Search
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Modern Search Engine Results Pages (SERPs) present complex layouts where multiple elements compete for visibility. Attention modelling is crucial for optimising web design and computational advertising, whereas attention metrics can inform ad placement and revenue strategies. We introduce AdSight, a method leveraging mouse cursor trajectories to quantify in a scalable and accurate manner user attention in multi-slot environments like SERPs. AdSight uses a novel Transformer-based sequence-to-sequence architecture where the encoder processes cursor trajectory embeddings, and the decoder incorporates slot-specific features, enabling robust attention prediction across various SERP layouts. We evaluate our approach on two Machine Learning tasks: (1)~emph{regression}, to predict fixation times and counts; and (2)~emph{classification}, to determine some slot types were noticed. Our findings demonstrate the model's ability to predict attention with unprecedented precision, offering actionable insights for researchers and practitioners.

Problem

• Quantify user attention in multi-slot search results
• Predict fixation times and counts accurately
• Determine noticed slot types in SERP layouts

Innovation

• Transformer-based sequence-to-sequence architecture for attention prediction
• Mouse cursor trajectories to quantify user attention
• Scalable and accurate attention modeling in SERPs