Skip to content

RECAPS Home

Satin Wuker edited this page Mar 12, 2024 · 1 revision

Welcome to the RECAPS wiki!

RECAPS Wiki

Welcome to the RECAPS Wiki! This is a comprehensive resource for understanding and using the RECAPS (Research Exploration and Collaboration Agent for Proposition Synthesis) platform. Here, you'll find detailed documentation, guides, and examples to help you get started and make the most out of this powerful research tool.

Table of Contents

  1. Introduction
  2. Installation
  3. Usage
  1. API Reference
  2. Contributing
  3. License

Introduction

RECAPS is a cutting-edge research copilot designed to revolutionize the way researchers approach complex questions and collaborate on projects. It leverages Large Language Models, advanced algorithms, prompting techniques, and natural language processing techniques to generate a comprehensive and intuitive representation of research findings.

Installation

To install RECAPS, follow these steps:

  1. Clone the repository: git clone https://github.com/ReLink-Inc/RECAPS.git
  2. Install the required dependencies: pip install -r requirements.txt

Usage

Inputting Research Questions

To begin, you'll need to input your research question into the RECAPS interface. The system will then break down the question into smaller, manageable sub-questions.

Recursive Question Subdivision

RECAPS intelligently breaks down research questions into smaller, manageable sub-questions, allowing for a more focused and efficient exploration of the problem space.

Candidate Solution Proposal

By analyzing the sub-questions, RECAPS proposes potential solutions, helping researchers identify promising avenues for further investigation.

Directed Acyclic Graph (DAG) Representation

The system organizes the propositions into a DAG, where each node represents an answer to a sub-research question. This visual representation provides a clear overview of the research landscape and helps identify connections between different aspects of the problem.

Bidirectional Graph Building

RECAPS can construct the DAG from both ends, either by recursively subdividing the research question or by cumulatively building propositions from the ground up. This flexibility allows researchers to approach the problem from multiple angles and gain a comprehensive understanding of the subject matter.

Document-Based Proposition Synthesis

Given a set of documents relevant to the research question, RECAPS can extract and synthesize information to generate propositions. This feature enables researchers to leverage existing knowledge and incorporate it seamlessly into their research.

Collaborative Research Platform

The DAG representation serves as a foundation for effective collaboration among researchers. By sharing and contributing to the same project, team members can work together to refine propositions, identify gaps in knowledge, and ultimately reach a well-supported conclusion.

API Reference

For detailed information on the RECAPS API and its various endpoints, please refer to the API Reference.

Contributing

We welcome contributions from the research community to enhance RECAPS and make it an even more powerful tool for research exploration and collaboration. If you'd like to contribute, please follow our contribution guidelines and submit a pull request.

License

RECAPS is released under the MIT License.