hongleizhang/RSPapers

What it solves

Researchers and practitioners often struggle to keep up with the fast-moving field of recommender systems. This project organizes the essential literature into one place, covering classic methods, deep learning approaches, social and knowledge-graph based methods, cold-start, conversational, industrial, privacy, large language model (LLM), and agentic recommender systems. It also collects tutorials from top conferences and comprehensive surveys, so readers can quickly find authoritative entry points instead of searching scattered sources.

How it works

The repository is a structured, hand-curated list of papers and tutorials. Papers are grouped into 17 numbered categories, such as General RS, Social RS, Deep Learning-based RS, CTR Prediction, Knowledge Graph for RS, LLM for RS, and Agentic RS. Each entry follows a simple format: author name, paper title, and venue/year. The README also links to tutorials and surveys at the top, and the maintainers update the list weekly. New subtopics (like Agentic RS and LLM for RS) have been recently added, and contributors can submit new papers via pull requests using a provided template.

Who it's for

This is aimed at researchers, graduate students, and industry engineers who study or build recommender systems. It is useful for anyone looking for a comprehensive bibliography to start learning a subtopic, find survey papers, or track recent trends such as LLM-based and agentic recommenders.

Highlights

• Covers 17 distinct subtopics, from classic collaborative filtering to agentic and LLM-based recommender systems
• Includes high-quality tutorials from top conferences like RecSys, ICML, and WWW
• Contains a large collection of surveys summarizing decades of research
• Updated weekly, with clear labels for new additions and easy contribution via pull requests