Modern information retrieval (IR) is built on pretrained language models whose parametric knowledge has driven strong gains in semantic understanding. Yet, this knowledge is frozen at training time, opaque, and shallow on long-tail, domain-specific, and fast-changing information, yielding imprecise, non-factual, or poorly grounded results. Knowledge-enhanced IR addresses this by augmenting retrieval systems—and the language models that power them—with external knowledge: unstructured or structured, gold or synthetic, increasingly across modalities. Large language models are now reshaping how this knowledge is acquired and integrated—for example, autonomous agents that plan, reason, and decide what to search; automated knowledge curation and linking; and systems that interleave retrieval and generation, learning when to consult external evidence and when to answer from parametric knowledge.

The Knowledge-Enhanced Information Retrieval (KEIR) workshop brings together researchers from Information Retrieval, NLP, Data Mining, and Knowledge Management to advance the principled integration of external knowledge into retrieval systems and the language models that power them.

Submission Guidelines

We invite both short and long papers presenting original unpublished work, work-in-progress research, or suitable CIKM 2026 rejected papers, as well as position, applied, and resource papers. All papers must be original and not simultaneously submitted to another journal or conference. We especially encourage submissions addressing knowledge-intensive and domain-sensitive applications where factuality, trustworthiness, and discovery are essential, including medicine, law, finance, and science.

List of Topics

• Knowledge-enhanced retrieval and recommendation models, both representational and generative
• Knowledge-aware data augmentation and query processing
• Knowledge-enhanced agentic IR: LLM agents that plan, reason, and decide when, what, and how to retrieve, navigate structured knowledge, and curate and link knowledge
• Self-evolving, dynamic knowledge and automated discovery
• Multimodal knowledge sources for IR
• Test-time adaptation of retrievers and language models
• Efficient knowledge-enhanced retrieval
• Evaluation, interpretability, uncertainty quantification, bias, and fairness
• Domain-specific retrieval applications

Committees

Organizing Committee

• Giacomo Frisoni (University of Bologna)
• Luca Ragazzi (University of Bologna)
• Gianluca Moro (University of Bologna)
• Zaiqiao Meng (Huazhong University of Science and Technology)
• Jinyuan Fang (University of Glasgow)
• Yufang Hou (IT:U Austria & IBM Research)

Publication

Authors may optionally publish an extended version of their accepted paper in a Springer LNCS post-proceedings volume; participation is entirely optional, helping preserve the opportunity to submit a revised version of the work to another conference or journal if desired.

Venue

The conference will be held in Sapienza University of Rome, Italy, co-located with ACM CIKM 2026.

Contact

All questions about submissions should be emailed to keir.ai.workshop@gmail.com.