Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

InteractiveSurvey: An LLM-based Personalized and Interactive Survey Paper Generation System

Zhiyuan Wen, Jiannong Cao, Zian Wang, Beichen Guo, Ruosong Yang, Shuaiqi Liu

TL;DR

The paper addresses the challenge of producing high-quality survey papers amid rapidly expanding literature. It introduces InteractiveSurvey, an LLM-based interactive system that combines automatic reference searching, MinerU-based parsing, HyDE-informed categorization, semantic clustering, and an interactive UI to iteratively refine outlines and content, resulting in structured, multimodal surveys with figures and citations. Key contributions include the first interactive survey-generation platform with modifiable intermediate outputs, open-source deployment, and demonstrated improvements in content quality and usability compared with mainstream LLMs and SOTA methods. Findings show InteractiveSurvey achieves superior coverage, structure, and relevance, completes a full survey in about 35 minutes on moderate hardware, and attains a high usability score (SUS 84.4/100), underscoring its practical impact for researchers and developers. The work paves the way for faster, tailored survey production and invites further enhancements such as multilingual capabilities and broader deployment scenarios.

Abstract

The exponential growth of academic literature creates urgent demands for comprehensive survey papers, yet manual writing remains time-consuming and labor-intensive. Recent advances in large language models (LLMs) and retrieval-augmented generation (RAG) facilitate studies in synthesizing survey papers from multiple references, but most existing works restrict users to title-only inputs and fixed outputs, neglecting the personalized process of survey paper writing. In this paper, we introduce InteractiveSurvey - an LLM-based personalized and interactive survey paper generation system. InteractiveSurvey can generate structured, multi-modal survey papers with reference categorizations from multiple reference papers through both online retrieval and user uploads. More importantly, users can customize and refine intermediate components continuously during generation, including reference categorization, outline, and survey content through an intuitive interface. Evaluations of content quality, time efficiency, and user studies show that InteractiveSurvey is an easy-to-use survey generation system that outperforms most LLMs and existing methods in output content quality while remaining highly time-efficient.

InteractiveSurvey: An LLM-based Personalized and Interactive Survey Paper Generation System

TL;DR

The paper addresses the challenge of producing high-quality survey papers amid rapidly expanding literature. It introduces InteractiveSurvey, an LLM-based interactive system that combines automatic reference searching, MinerU-based parsing, HyDE-informed categorization, semantic clustering, and an interactive UI to iteratively refine outlines and content, resulting in structured, multimodal surveys with figures and citations. Key contributions include the first interactive survey-generation platform with modifiable intermediate outputs, open-source deployment, and demonstrated improvements in content quality and usability compared with mainstream LLMs and SOTA methods. Findings show InteractiveSurvey achieves superior coverage, structure, and relevance, completes a full survey in about 35 minutes on moderate hardware, and attains a high usability score (SUS 84.4/100), underscoring its practical impact for researchers and developers. The work paves the way for faster, tailored survey production and invites further enhancements such as multilingual capabilities and broader deployment scenarios.

Abstract

The exponential growth of academic literature creates urgent demands for comprehensive survey papers, yet manual writing remains time-consuming and labor-intensive. Recent advances in large language models (LLMs) and retrieval-augmented generation (RAG) facilitate studies in synthesizing survey papers from multiple references, but most existing works restrict users to title-only inputs and fixed outputs, neglecting the personalized process of survey paper writing. In this paper, we introduce InteractiveSurvey - an LLM-based personalized and interactive survey paper generation system. InteractiveSurvey can generate structured, multi-modal survey papers with reference categorizations from multiple reference papers through both online retrieval and user uploads. More importantly, users can customize and refine intermediate components continuously during generation, including reference categorization, outline, and survey content through an intuitive interface. Evaluations of content quality, time efficiency, and user studies show that InteractiveSurvey is an easy-to-use survey generation system that outperforms most LLMs and existing methods in output content quality while remaining highly time-efficient.

Paper Structure

This paper contains 26 sections, 3 equations, 5 figures, 5 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Literature Review Generation
  4. Automated Survey Systems
  5. InteractiveSurvey
  6. Automatic Reference Searching
  7. Reference Parsing
  8. Personalized Reference Categorization
  9. Categorization Context Retrieval
  10. Survey Outline Generation
  11. Survey Content Generation
  12. Evaluation
  13. Content Quality
  14. Time Efficiency
  15. User Study with SUS Test
  16. ...and 11 more sections

Figures (5)

  • Figure 1: Comparison of the number of all research papers and survey papers released on arXiv.org over the past 10 years (2015–2024).
  • Figure 2: An Overview of InteractiveSurvey. Steps 2,4,5, and 6 in user interactions are optional.
  • Figure 3: The HyDE Process for Retrieving Research Method Descriptions from References.
  • Figure 4: The pre-defined section names and the categorization names are as section titles, sub-section titles are generated by the LLM.
  • Figure 5: Time cost distribution in InteractiveSurvey