-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
Overview
Develop a web-based chatbot that translates natural language questions into Cypher queries to interact with a Neo4j database. The project involves integrating a lightweight Hugging Face LLM model, setting up a Neo4j database locally, building a user-friendly interface, and sharing the complete code on GitHub with thorough documentation.
Task Breakdown
1. LLM Model Integration
- Objective: Select and download a lightweight, efficient LLM model from Hugging Face.
- Steps:
- Research and identify an appropriate model capable of translating natural language into Cypher queries.
- Download and configure the model for local use.
- Optimize for performance and resource efficiency.
2. Neo4j Database Setup
- Objective: Install and configure Neo4j locally.
- Steps:
- Install Neo4j on your local machine.
- Create or select a test database for querying.
- Ensure connectivity and proper functioning of the database.
3. Web Interface Development
- Objective: Build a simple web interface for user input and output display.
- Steps:
- Create a web page where users can enter natural language queries.
- Integrate the LLM model to convert these queries into Cypher queries.
- Execute the generated Cypher queries on the Neo4j database.
- Display the query results in a clear, accessible format.
4. Project Sharing and Documentation
- Objective: Share the project on GitHub with complete instructions.
- Steps:
- Upload all project code to a GitHub repository.
- Provide detailed documentation that includes:
- Setup instructions for both the LLM model and Neo4j.
- Steps to run the web interface.
- Explanation of how natural language is translated into Cypher queries.
- List of dependencies and environment configurations.
Deliverables
- A fully functional web-based chatbot that:
- Accepts natural language questions.
- Translates them into Cypher queries using the selected Hugging Face LLM model.
- Executes these queries against a local Neo4j database.
- Displays the results to the user.
- A GitHub repository containing:
- All project code.
- Comprehensive documentation to reproduce the setup and execution.
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels