Zero-shot Document Retrieval with Hybrid Pseudo-document Retriever (ICASSP 2025)

Metadata

| Field | Content | |-------|---------| | Title | Zero-shot Document Retrieval with Hybrid Pseudo-document Retriever, ICASSP 2025 | | Source | IEEE Xplore | | Type | Academic paper | | Section | Unclassified |

One-line Summary

This work targets zero-shot document retrieval using a hybrid pseudo-document retriever, addressing efficiency, scalability, and user-intent understanding in large-scale search and recommendation systems.

Background and Motivation

In large-scale search, recommendation, and personalization systems, information retrieval has long faced challenges in efficiency, scalability, and user-intent understanding. Traditional pipelined approaches tend to separate retrieval, ranking, and generation, making it hard to meet modern demands for natural-language interaction, multi-hop reasoning, and real-time knowledge in the LLM era. This paper is proposed against this backdrop, aiming to advance the theory and practice at this intersection.

The core scenarios include open-domain information access, enterprise knowledge retrieval, conversational search, semantic understanding in recommendation, and end-to-end architectures that coordinate external knowledge sources with generative models.

Core Contributions

• A unified perspective that brings scattered related work into a comparable framework.
• A clear decomposition of method components (representation learning, retriever, re-ranker, planner, generator, feedback mechanisms) to facilitate engineering adoption.
• Reproducible benchmarks, datasets, or taxonomies that lower the entry cost for follow-up researchers.
• Discussion of interfaces with emerging paradigms such as LLM tool calling, reinforcement learning, and multi-agent collaboration.
• Explicitly listed open problems: evaluation trustworthiness, latency and cost, hallucination and safety, and cross-lingual / multimodal extension.

Method / System Architecture

The typical pipeline follows: problem formulation → model/system design → training or construction → inference:

1. Input and representation: encode queries, documents, and user context into dense or sparse representations, or structured prompts; 2. Core modules: retriever, re-ranker, planner, memory, tool interfaces, chained or parallelized by task; 3. Learning strategies: supervised fine-tuning, contrastive learning, distillation, RL (including process rewards), bootstrapped data synthesis; 4. Inference strategies: single-pass retrieval, iterative retrieval, parallel sub-queries, early stopping and budget control.

> Note: the original abstract text was not fully reproduced in the source post; consult the IEEE page for exact details.

Experiments and Evaluation

Typical evaluation setups in this area include:

• Datasets: MS MARCO, BEIR, Natural Questions, domain-specific corpora;
• Metrics: nDCG@10, MRR, Recall@k, Hit@k, human preference, task success rate, latency and token cost;
• Baselines: BM25, dense retrieval, cross-encoder re-ranking, retrieval-free LLMs, commercial search APIs;
• Ablations: contribution of retrieval steps, re-ranking depth, and training data scale.

Specific numeric results should be verified against the original PDF tables.

Key Takeaways

1. Architecture: cascaded retrieve–rank–generate remains mainstream, but agentic paradigms are making "whether and how many times to retrieve" itself learnable; 2. Data: high-quality instruction data and click/session logs matter equally; synthetic data must guard against leakage and distribution shift; 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM-as-judge needs cross-validation with human evaluation; 4. Product: latency, cost, explainability, and safety are hard constraints for industrial deployment.

Limitations and Future Work

Potential limitations include experiment scale constrained by GPU budgets, benchmarks misaligned with real user distributions, unknown cross-lingual generalization from English-centric data, and safety risks of agents on the open web. Future directions include more efficient test-time compute allocation, deeper integration with knowledge graphs and structured databases, and causal/fairness constraints for recommender systems.

Cross-references

• A Comprehensive Study of Knowledge Editing for Large Language Models (arXiv:2401.01286)
• INTERS: Unlocking the Power of Large Language Models in Search (arXiv:2401.06532)
• Recommendation as Instruction Following (ACM DOI: 10.1145/3708882)
• RouteLLM: Learning to Route LLMs with Preference Data (arXiv:2406.18665)
• Translational Generative Retrieval via Potential Query Generation (ICASSP)
• Representation Learning with Large Language Models for Recommendation (ACM DOI: 10.1145/3589334.3645458)

Glossary

| Term | Meaning | |------|---------| | IR | Information Retrieval | | RAG | Retrieval-Augmented Generation | | LTR | Learning to Rank | | nDCG | Normalized Discounted Cumulative Gain | | Agentic Search | Modeling search as sequential decision-making and tool calling | | Gen-IR | Generative Information Retrieval |

References

• Original paper: Zero-shot Document Retrieval with Hybrid Pseudo-document Retriever, ICASSP 2025 — IEEE Xplore