A Survey of Conversational Search (ACM, September 2025)

Source: https://dl.acm.org/doi/full/10.1145/3759453

Key points

• This is a peer-reviewed ACM survey (September 2025, DOI: 10.1145/3759453) covering conversational search: information access through multi-turn natural-language dialogue.
• It reviews the evolution of the field from neural retrieval foundations (BERT reranking, dense passage retrieval / DPR) to RAG-based systems and modern agentic search.

Research background and motivation

Large-scale search, recommendation, and personalization systems have long faced challenges around efficiency, scalability, and user intent understanding. Traditional pipeline approaches treat retrieval, ranking, and generation as disjoint stages, which struggles to meet the LLM-era demand for natural-language interaction, multi-hop reasoning, and access to fresh knowledge. This survey aims to systematically organize the theory and practice at this intersection.

Core scenarios include:

• Open-domain information access
• Enterprise knowledge retrieval
• Conversational search and multi-turn query reformulation
• Semantic understanding in recommendation
• End-to-end architectures combining external knowledge sources with generative models

Taxonomy of methods

| Dimension | Sub-classes | Representative ideas | Strengths | Limitations | |---|---|---|---|---| | Modeling paradigm | Discriminative / generative retrieval | Bi-encoders, cross-encoders, DSI | Mature, scalable | Semantic drift, update cost | | LLM integration | RAG / Agent / Tool-use | Retrieval augmentation, search agents, API calls | Flexible, interpretable | Latency, error propagation | | Optimization objectives | Relevance / diversity / freshness | Multi-objective LTR, RLHF, online learning | Business-aligned | Scarce annotations | | Evaluation | Offline / online / human | nDCG, MRR, LLM-as-judge, A/B tests | Comparable | Deviates from true satisfaction |

Four main research lines

The survey contrasts four major lines of work in neural information retrieval:

• Dense retrieval: high recall and low latency; well suited for first-stage retrieval, but sensitive to domain shift and long-tail queries.
• Late interaction (e.g., ColBERT): higher precision, at the cost of larger indexes.
• Generative IR: directly "generates" documents via tokens or docids, simplifying the cascade but complicating index updates.
• Agentic search: models search as sequential decision-making, supporting multi-hop reasoning and self-reflection.

Timeline of the field

• 2019–2021: BERT reranking and DPR establish neural retrieval foundations.
• 2022–2023: RAG and FreshLLMs drive retrieval-generation fusion.
• 2024 onward: conversational/agentic search and generative recommender systems (Gen-RecSys) expand rapidly.
• 2025–2026: RL-trained search agents, Deep Research, and GraphRAG emerge as new growth areas.

Evaluation paradigms

• Datasets: MS MARCO, BEIR, Natural Questions, domain-specific corpora
• Metrics: nDCG@10, MRR, Recall@k, Hit@k, human preference, task success rate, latency, token cost
• Baselines: BM25, dense retrieval, cross-encoder reranking, no-retrieval LLMs, commercial search APIs
• Ablations: contribution of retrieval steps, reranking depth, and training data scale

Conclusions and insights

1. Architecture: cascade retrieval + reranking + generation remains mainstream, but agentic paradigms make the number and policy of retrieval steps itself learnable. 2. Data: high-quality instruction data and click/session logs are both critical; synthetic data risks knowledge leakage and distribution shift. 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM-as-judge should be cross-validated with human assessment. 4. Deployment: latency, cost, interpretability, and safety are hard constraints for industrial systems, not just academic benchmarks.

Open problems and future directions

Noted gaps include the lack of unified benchmarks, limited reproducibility on private data, biases in LLM-based evaluation, and the safety/cost constraints of agentic systems. Future directions include finer-grained process supervision, joint retrieval-reasoning training, enterprise metadata governance, and multimodal/cross-lingual consistency.

Related entries

• Engineering Conversational Search Systems: A Review of Applications
• ChatRetriever: Adapting Large Language Models for Generalized and Robust Retrieval
• CoSearchAgent: A Lightweight Collaborative Search Agent
• ConvGQR: Generative Query Reformulation for Conversational Search

Glossary

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

> Note: Quantitative claims should be verified against the original PDF at the ACM Digital Library link above; this report is based on the survey's abstract and public metadata.