Content Recommendation Powered by Language Models

Dear Colleagues,

Content recommendation systems (RSs) are experiencing a profound transformation driven by the rapid emergence of Large Language Models (LLMs). Classical paradigms—rooted in collaborative filtering, content-based approaches, and matrix factorization—have long served as the backbone of personalized recommenders. However, these traditional methods often fall short when confronted with sparse data, the cold-start problem, evolving and nuanced user intent, and the need for deep semantic understanding. In contrast, the sophisticated understanding of the natural language of LLMs introduces unprecedented capabilities for modeling context, reasoning over user preferences, and generating adaptive, conversational recommendations.

By harnessing techniques such as instruction following, semantic embedding extraction, tool-augmented retrieval, and generative reranking, LLM-enhanced RSs can move beyond deterministic matching to deliver richer, more coherent, and more explainable user experiences. Their ability to integrate heterogeneous data sources, interpret multimodal content, and interact through natural language paves the way for next-generation systems capable of supporting domains as diverse as digital media, e-commerce, education, scientific discovery, or decision support.

This Special Issue seeks to explore emerging theories, models, and applications at the intersection of LLMs and recommendation technologies. We encourage contributions that advance our understanding of how LLMs can serve as feature extractors, generative engines, or end-to-end recommendation backbones, as well as papers that critically examine their limitations—including computational scalability, reliability, privacy, and ethical risks.

Topics of interest include, but are not limited to, the following:

• Generative and LLM-centric architectures for recommendation;
• Prompting, fine-tuning, and alignment strategies for personalized discovery;
• Conversational and agent-based recommendation interfaces;
• Hybrid systems combining LLMs with knowledge graphs or traditional RS models;
• Evaluation methodologies and explainability techniques for LLM-driven RSs;
• LLM-based zero-shot and few-shot solutions to the cold-start problem;
• Fairness, transparency, robustness, and responsible AI in recommendation;
• Efficient inference, distillation, and deployment for real-time scenarios.

We look forward to receiving high-quality submissions that contribute to shaping the future of intelligent, trustworthy, and human-centric recommendation systems.

Prof. Dr. Alberto Gil Solla
Prof. Dr. Yolanda Blanco Fernández
Prof. Dr. José Carlos López Ardao
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-anonymized peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• LLMs for recommendation
• LLM-based personalization
• semantic representation learning
• item and user embeddings
• prompt engineering and alignment
• retrieval-augmented recommendation (RAG)
• cold-start problem mitigation
• explainable and transparent recommendations
• scalable and efficient inference for RS
• multimodal content recommendation