AI4UAI4U
STORM

STORM

3
Visit
An LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations.

STORM Overview

STORM is an LLM-powered knowledge curation system designed to streamline knowledge acquisition by researching topics and generating full-length reports with citations.

How it Works

  • Research and Report Generation: The STORM system conducts internet-based research and generates a full-length report with citations.
  • Core Strategy: STORM automates the research process by asking high-quality questions, employing "Perspective-Guided Question Asking" and "Simulated Conversation" strategies.

Key Features

  • Perspective-Guided Question Asking: Analyzes existing articles to ask questions from different perspectives, providing a more comprehensive understanding of the topic.
  • Simulated Conversation: Simulates a conversation between a Wikipedia writer and a topic expert to update the understanding of the topic and ask follow-up questions.

Co-STORM

Co-STORM enhances STORM's capabilities by enabling human-machine collaboration for more aligned and optimized information retrieval and knowledge curation.

  • Collaboration Protocol: Co-STORM adopts a collaborative protocol to support smooth collaboration among LLM experts, moderators, and human users.
  • Dynamic Mind Map: Co-STORM maintains a dynamically updated mind map to organize collected information, building a shared conceptual space between the human user and the system.

Summary

STORM and Co-STORM aim to simplify knowledge acquisition and report generation processes, improving research efficiency and quality through LLM-driven systems.