Skip to content

Latest commit

 

History

5 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NEXUS: Personal Knowledge Base

NEXUS is a high-speed, hybrid query system that transforms your unstructured text documents into a fully queryable, visual knowledge graph.

It uses a Large Language Model (Google's Gemini API) for its world-class accuracy in data extraction and a hyper-fast, custom-built C-Core Engine for instantaneous graph queries.

This "best of both worlds" architecture gives you the intelligence of a massive AI model and the sub-millisecond query speed of pure C.

🏛️ Architecture

NEXUS is built in two distinct phases:

  1. Phase 1: Ingestion (Slow, Smart) When you upload a document, the text is sent to the Google Gemini API. The AI reads the text, extracts all entities and relationships, and formats them into a simple graph_data.txt file. This step is "slow" (it can take 30-60 seconds) but only happens once.

    Document.txt --> [Python Backend] --> [Google Gemini API] --> graph_data.txt

  2. Phase 2: Querying (Instant, Fast) All subsequent queries use the custom C-Core engine. This compiled program loads the graph_data.txt file into memory and uses highly-optimized C data structures to perform searches. This step is "fast," with most queries returning in less than 1ms.

    User Query --> [Python Backend] --> [C-Core Engine] --> Instant Results


✨ Features

NEXUS can perform 5 types of advanced queries on your documents:

  • Path Finding: Finds the shortest path between two concepts (e.g., "Stack" to "LIFO").
  • Topic Discovery: Uses Depth-First Search (DFS) to find all separate, disconnected clusters of information in your graph.
  • Visual Mind Map: Generates a visual, interactive graph of how a concept is connected, rendered using Mermaid.js.
  • Q&A: Lists all direct, 1-to-1 relationships for a given node.
  • Autocomplete: Provides instant search suggestions using a custom-built Trie data structure.

🛠️ Tech Stack

Component Technology Purpose
AI Extraction Google Gemini API World-class entity & relation extraction.
Backend Python Flask API server to handle requests.
Query Engine Pure C Graph algorithms & data structures (DSA).
Frontend HTML/CSS/JS User interface for uploading and querying.
Visualization Mermaid.js Renders the visual, interactive mind maps.

🚀 Quick Start

  1. Get Your API Key: This project requires a Google Gemini API key.
  2. Set Up the Project: Follow the SETUP.md file for detailed installation and compilation steps.
  3. Run the Server:
    # (From the 'backend' folder)
    python app.py
  4. Open the Frontend: Open the frontend/index.html file in your browser.

Usage

  1. Upload: Click "Process Document" and select a .txt file.
  2. Wait: Allow 30-60 seconds for the Gemini API to process the text (you only do this once per file).
  3. Query: Once the nodes appear in the "Available Nodes" list, use the query interface to explore your new knowledge graph.

About

No description, website, or topics provided.

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages