Knowledge Graph from Unstructure Data
Glen Newton, M. Shajith Revanth, P. Minish, Kuldeep Reddy · 2025
This project presents a client-side, knowledge graph system that dynamically extracts and visualizes semantic relationships from unstructured natural language input. Unlike traditional keyword-based methods, this system uses lightweight Natural Language Processing (NLP) to interpret the contextual meaning of user queries. Unlike traditional keyword-based methods, this system uses lightweight Natural Language Processing (NLP) to interpret the contextual meaning of user queries. It identifies key entities and their relationships through in-browser logic and parsing, transforming them into nodes and edges rendered instantly as a knowledge graph. Built with React, TypeScript (TSX), and ReactFlow, the interface offers an intuitive experience for exploring semantic structures without relying on any backend or external database. This fully browser-based architecture ensures fast, private, and responsive interaction. The system is well- suited for applications such as semantic search, concept discovery, educational tools, and interactive data exploration—enabling users to better understand and navigate the relationships embedded in text.