Bridge the Gap between Past and Future: Siamese Model Optimization for Context-Aware Document Ranking

2024-10-21

International Conference on Information and Knowledge Management

AI Summary

To address the limited document ranking accuracy in multi-turn search sessions caused by dynamic user intent evolution, this paper proposes ForeRanker—a novel dual-branch Siamese ranking framework incorporating future-context modeling. ForeRanker jointly models historical session context and future query behavior in parallel, mitigating historical–future model inconsistency via mutual supervision, pseudo-label co-training, and peer knowledge distillation; it further introduces a dynamic gating mechanism for cross-temporal intent alignment. Crucially, ForeRanker implicitly captures future intent during inference without requiring actual future queries. Evaluated on standard benchmarks including TREC CAR and MSDialog, ForeRanker achieves an average 3.2% improvement in nDCG@5 over state-of-the-art methods, demonstrating the efficacy of future-context modeling for conversational retrieval.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Learning Preferences or Rankings
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

In the realm of information retrieval, users often engage in multi-turn interactions with search engines to acquire information, leading to the formation of sequences of user feedback behaviors. Leveraging the session context has proven to be beneficial for inferring user search intent and document ranking. A multitude of approaches have been proposed to exploit in-session context for improved document ranking. Despite these advances, the limitation of historical session data for capturing evolving user intent remains a challenge. In this work, we explore the integration of future contextual information into the session context to enhance document ranking. We present the siamese model optimization framework, comprising a history-conditioned model and a future-aware model. The former processes only the historical behavior sequence, while the latter integrates both historical and anticipated future behaviors. Both models are trained collaboratively using the supervised labels and pseudo labels predicted by the other. The history-conditioned model, referred to as ForeRanker, progressively learns future-relevant information to enhance ranking, while it singly uses historical session at inference time. To mitigate inconsistencies during training, we introduce the peer knowledge distillation method with a dynamic gating mechanism, allowing models to selectively incorporate contextual information. Experimental results on benchmark datasets demonstrate the effectiveness of our ForeRanker, showcasing its superior performance compared to existing methods.

Problem

• Enhance document ranking using future contextual information
• Optimize siamese model for context-aware document ranking
• Address evolving user intent with session context integration

Innovation

• Siamese model with history and future context
• Peer knowledge distillation with dynamic gating
• ForeRanker learns future-relevant information