
IterCQR iteratively trains a conversational query reformulation model without human-annotated rewrites, utilizing retrieval signals as a reward to generate queries optimized for off-the-shelf retrievers.


coremsg

iterative-conversational-query-reformulation-with-retrieval-guidance


Iterative Conversational Query Reformulation with Retrieval Guidance


title_rewrite


Conversational Query Reformulation (CQR) can be improved by leveraging key information extracted from initially retrieved documents to generate more retriever-friendly queries.


leveraging-retrieved-documents-to-enhance-conversational-query-reformulation-for-improved-search-performance


Leveraging Retrieved Documents to Enhance Conversational Query Reformulation for Improved Search Performance
