The field of information retrieval is witnessing significant advancements in query expansion and conversational search. Researchers are exploring innovative methods to improve retrieval performance by enriching queries with related terms and leveraging large language models. A key direction in this area is the development of efficient and effective query expansion techniques that balance relevance and diversity. Another notable trend is the use of iterative clarification and rewriting frameworks to improve conversational search. Noteworthy papers in this area include:

• Candidate Token Query Expansion, which achieves strong retrieval performance with significantly lower cost by leveraging unselected candidate tokens from large language models.
• ICR, which proposes an iterative rewriting scheme that pivots on clarification questions to improve retrieval performance.
• LESER, which fine-tunes a context-aware large language model using real-time search engine feedback to produce high-quality query expansions.
• AudioBoost, which generates synthetic queries conditioned on audiobook metadata to improve retrievability in Spotify's Search.