The EARL Workshop welcomes submissions on (1) emerging techniques for LLM-based RSs (including retrieval-augmented generation, multi-modal recommendation, reinforcement learning with human feedback, scalable fine-tuning methods, dynamic prompting strategies, and personalized conversational agents), (2) real-world applications of existing LLMs, and (3) critical challenges in ensuring the trustworthiness and responsibility of LLM-driven RSs.
In detail, topics of interest include, but are not limited to, the following:
- Integrating LLMs to enhance RSs.
- Leveraging LLM-generated data to improve traditional RSs.
- LLM fine-tuning or prompt engineering techniques for RSs.
- Developing interactive and conversational RSs with LLMs.
- Integration of reinforcement learning with LLMs to adapt recommendations based on user feedback.
- Leveraging retrieval-augmented generation (RAG) to improve relevance and diversity in recommendations.
- Multi-modal RSs powered by LLMs, e.g., techniques integrating text, images, and audio data.
- Few-shot and zero-shot learning for LLM-based RSs.
- Personalization strategies for LLM-powered RSs, including dynamic user modeling and real-time adaptation.
- Cross-domain and cross-lingual RSs utilizing multilingual and generalist LLMs.
- Applications of LLM-enhanced RSs in domains such as finance, streaming platforms, and social networks.
- Scalability challenges in deploying LLM-powered RSs.
- Efficiency challenges in deploying LLM-powered RSs.
- Evaluation of LLM-powered RSs using novel metrics/standards.
- Evaluation of LLM-powered RSs using human feedback.
- Enhancing transparency in LLM-powered RSs.
- Enhancing fairness in LLM-powered RSs.
- Enhancing explainability in LLM-powered RSs.
- Trustworthy recommendation with LLMs, addressing bias, safety, privacy, and authenticity issues.
- Responsible AI practices in LLM-powered RSs, emphasizing ethical considerations and sustainable AI.
Paper submission
-
Authors are encouraged to submit contributions in one of the following formats.
-
This workshop is non-archival, meaning that submissions will neither be indexed nor have formal proceedings.
- Accepted papers will appear on the workshop website.
- We welcome submissions that are currently under review at other venues, provided this does not breach the dual-submission or anonymity policies of those venues.
-
The review process will be double-blind.
Important Dates:
- Call for Papers publication: April 25, 2025
- Paper submission deadline: July 10, 2025
- Reviewer deadline: July 31, 2025
- Author notification: August 6, 2025
- Camera-ready version deadline: August 20, 2025