StockGenie: AI-Driven Stock Market Assistant and Forecasting System

Abstract

Stock market investing demands constant evaluation
of extensive financial datasets, accurate recognition of market
trends, and careful management of investment risks. These
requirements often create significant challenges for beginner
investors who may lack analytical expertise and access to
advanced decision-support tools. While many existing trading
platforms provide real-time market data, they frequently fail to
offer intelligent forecasting mechanisms and structured learning
environments that aid users in understanding market dynamics.
This paper proposes StockGenie, an AI-driven stock market
assistance system developed to facilitate informed investment
decisions through predictive modeling, visual analytics, portfolio
evaluation, and simulated trading experiences. The system utilizes
time-series forecasting approaches, including Autoregressive
Integrated Moving Average (ARIMA) and Long Short-Term
Memory (LSTM) neural networks, to examine historical stock
market data and generate future price predictions. In addition
to forecasting capabilities, StockGenie incorporates an interactive
visualization dashboard, a portfolio advisory component for
risk-aware investment analysis, a virtual trading simulator for
hands-on practice, and an AI-powered chatbot that provides
instant guidance and explanations related to market behavior.
Experimental evaluation using historical stock market datasets
was conducted to assess forecasting performance. Quantitative
metrics including Root Mean Square Error (RMSE) and Mean
Absolute Error (MAE) were used to compare ARIMA and LSTM
models. The results indicate that the LSTM model achieves
lower prediction error and improved trend prediction capability
compared to traditional statistical approaches