This repository serves as a curated bibliography of academic papers and resources focused on recommendation systems. It is designed for researchers, engineers, and practitioners seeking to quickly access foundational and state-of-the-art literature in the field, offering a structured overview of key concepts and advancements.

How It Works

The project functions as an organized index rather than executable code. It meticulously categorizes seminal and recent research papers across various recommendation system sub-domains, including retrieval, deep learning architectures, embedding techniques, LLM applications, and industry case studies. This structured approach facilitates efficient navigation and discovery of relevant academic contributions.

Quick Start & Requirements

No installation or code execution is required. Users need to independently access the papers referenced, typically through academic search engines or institutional subscriptions. A foundational understanding of machine learning and recommendation system principles is essential to effectively utilize the curated list.

Highlighted Details

• Comprehensive coverage spans from classic Collaborative Filtering (CF) to advanced LLM-based recommendation paradigms.
• Detailed categorization includes deep learning models (e.g., Wide&Deep, DeepFM, DIN), embedding methods (e.g., Word2Vec, Node2vec), LLM applications, and industry case studies from major tech companies.
• Dedicated sections address critical areas such as evaluation methodologies, reinforcement learning applications, and challenges like cold start and bias mitigation.

Maintenance & Community

The repository is indicated to be dynamically updated. The maintainer, Wang Zhe, provides contact information (Email: wzhe06@gmail.com, LinkedIn, Zhihu) for discussions and inquiries, suggesting ongoing curation and engagement.

Licensing & Compatibility

No specific license is stated for the repository itself. The content comprises references to academic papers, each subject to its own copyright and licensing terms. Commercial use or integration would depend on the licensing of the individual papers.

Limitations & Caveats

This is a curated list of papers, not a software project with runnable code. Users must source and access the papers independently. The depth and recency of coverage within each category are dependent on the maintainer's ongoing efforts.