Knowledge-Enhanced Information Retrieval

KEIR 2025

April 10, 2025 @ ECIR
Lucca, Italy

Abstract

Pretrained language models (PLMs) like BERT and GPT-4 have become foundational to modern information retrieval (IR) systems. However, existing PLM-based IR models primarily rely on knowledge learned during training for predictions, limiting their ability to access and incorporate external, up-to-date, or domain-specific information. Consequently, current IR systems struggle with semantic nuances, contextual relevance, and domain-specific challenges.

This workshop (KEIR @ ECIR 2025) serves as a platform to discuss innovative approaches that integrate external knowledge, aiming to enhance the effectiveness of information retrieval in a rapidly evolving technological landscape. Our goal is to bring together researchers from academia and industry to explore various aspects of knowledge-enhanced information retrieval.

We invite researchers to submit their latest work to the KEIR @ ECIR 2025 workshop on various aspects of knowledge-enhanced information retrieval, including models, techniques, data collection, and evaluation methodologies. Topics covered will include, but are not limited to:

• Knowledge-enhanced information retrieval models
• Knowledge-enhanced approaches for query processing, including query parsing, query expansion, relevance feedback, and query reformulation
• Knowledge-enhanced recommendation models
• Knowledge-enhanced language models for retrieval
• Data augmentation for knowledge-enhanced information retrieval
• Large language model enhanced information retrieval
• Data collection for knowledge-enhanced information retrieval
• Knowledge-enhanced retrieval-augmented generation models
• Knowledge-aware fine-tuning and optimization methods for large language models tailored to IR systems
• Applications of knowledge-enhanced retrievals, such as dialogue systems, question answering, summarisation, and other domain-specific applications
• Evaluation methodologies for knowledge-enhanced retrieval
• The interpretability and analysis of knowledge-enhanced models for IR, including potential biases, fairness, and ethical considerations