Search for a command to run...
Abstract — The continuous expansion of the cryptocurrency market has intensified the demand for intelligent, trustworthy, and user-centric analytical platforms. Cryptosden is a comprehensive MERN-stack (MongoDB, Express.js, React, Node.js) web application engineered to overcome the shortcomings of existing cryptocurrency platforms by providing a centralized, secure, and highly informative hub for both novice and experienced users. The platform delivers dynamically updating crypto values sourced from the CoinGecko API, enabling users to monitor over 3,000 digital assets in real time across four fiat currencies — USD, INR, EUR, and GBP. Cryptosden distinguishes itself through four proprietary AI-powered modules: a composite Coin Trust Score that evaluates market health across six weighted data streams; an Emotional Volatility Index (EVI) that quantifies collective community sentiment; a Gated Recurrent Unit (GRU)- based Price Prediction Engine that forecasts coin prices over four time horizons using 13 engineered technical features; and a parallel Anomaly Detection and Whale Alert System employing Z-score statistics, Isolation Forest machine learning, and blockchain transaction monitoring. A Gemini-powered AI chatbot further assists users in interpreting complex market signals in natural language. The platform targets retail cryptocurrency participants who need consolidated analytical intelligence to make informed buy, sell, or hold decisions — without the system providing direct financial advice. This paper presents the full architecture, AI algorithms, experimental results, and comparative evaluation of Cryptosden. Index Terms — Cryptocurrency analytics, GRU price prediction, Emotional Volatility Index, Anomaly detection, Whale alerts, MERN stack, Trust Score, Isolation Forest, VADER NLP, Gemini AI chatbot.
Published in: INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT
Volume 10, Issue 04, pp. 1-9
DOI: 10.55041/ijsrem58845