SurveyX: Open-Source Automated Academic Research & Paper Generator 146

An advanced academic survey automation system that leverages large language models (LLMs) to generate high-quality, domain-specific academic papers and surveys.

#Ai工具箱 #Ai开源项目

Visit SurveyX: Open-Source Automated Academic Research & Paper Generator

SurveyX: Open-Source Automated Academic Research & Paper GeneratorIntroduction

What is SurveyX?

SurveyX is an advanced academic survey automation system jointly developed by Renmin University of China, the University of Sydney, and Northeastern University in China. It leverages large language models (LLMs) to generate high-quality, domain-specific academic papers and surveys. It aims to address the limitations of current LLM-based automated methods in content generation. Users only need to provide the titles and keywords of papers for literature retrieval, and can request comprehensive academic papers or surveys on specific topics.

SurveyX Key Features

Automatic Paper and Survey Generation: Simply provide it with the paper title and keywords, and it will generate a comprehensive academic paper or survey on your specified topic.
Reference Lookup: It can filter through a large volume of literature to identify references closely related to the topic, using a keyword expansion algorithm and a two-step filtering method.
Structured Information Processing: It organizes the key information from references into structured data, enriching the information and enabling the language model to perform more effectively.
Multiple Output Formats: In addition to text content, it can also generate charts and diagrams, making the survey results more intuitive to present.

SurveyX Usage Steps

Submit Requirements: You need to create a new issue and fill in the paper title, keywords, and optionally your email address.
Wait for Generation: After submission, the generated paper will be placed in your request folder, and it typically takes 1 - 2 business days to process.

SurveyX Technical Highlights

Decomposed Survey Workflow: The survey writing process is split into two stages—preparation and generation—fully leveraging the capabilities of large language models.
Diverse Presentation Formats: In addition to text, it can also generate charts, making the survey more vivid in presentation.
Outline Optimization Strategy: The outline is optimized through a "separation-recombination" approach, ensuring the generated survey outline is logically clear and free of redundant content.

SurveyX Performance Metrics

High Scores Across All Metrics: Scores in coverage, structure, and relevance reach 4.95, 4.91, and 4.94 (out of 5), comparable to those of human experts.
Excellent Citation Quality: The F1 score reaches 81.52, which is very close to the human expert benchmark of 81.83.