The scope of our project encompasses the development of a web application focused on stock visualization. Our dashboard will center around a subset of the major companies listed on the NASDAQ and DJI- 30 to provide a comprehensive yet contained dataset. Specifically, stocks such as Apple (AAPL), Microsoft, Alphabet, and Cisco will comprise the core research universe. This would allow users to pick heavily traded securities and understand different algorithms performances during various years and market climate. 

Read More

The application will contain the five main tabs, each offering distinct features and functionalities. The first tab will provide Stock visualizations along with latest news about that stock analysis, while the second tab will present algorithmic trading graphs and the third tab will have Deep Learning Models which users can play with. The fourth tab will contain FinLLM which is a LLama model. It will be providing insights about the chosen stocks. Finally, the fifth tab will contain a discussion forum where new investors can discuss various financial topics and learn from their peers. Through these tabs, users will gain valuable insights into stock market trends, enabling them to make informed investment decisions. 

The FinLLM tab will leverage natural language processing and GenAI techniques to analyze relevant news articles, social media posts, and other textual data related to stocks. This analysis will then be transformed into suggestions such as potential concerns, summary, positive developments and so on; providing users with a comprehensive overview of analysis, stock mentions, and other relevant information. By presenting these insights in a user-friendly format, the application aims to simplify the interpretation of textual data and facilitate more informed decision-making processes. 

The algorithmic trading graphs tab will focus on providing users with a range of customizable graphs and charts that depict key financial indicators and trading patterns. This would include deep learning models and well-known trading strategies. Additionally, the application will incorporate interactive features, allowing users to zoom in, zoom out, and toggle between different time-frames to gain a deeper understanding of the data. 

In terms of functionality, the web application will offer a user-friendly interface that allows for seamless navigation between the tabs. Users will have the ability to switch between FinLLM and visualizations and algorithmic trading graphs effortlessly, enabling them to explore different aspects of stock market information in a cohesive manner.

To ensure the accuracy and reliability of the data presented, the application will integrate with reputable financial data providers and news sources. Real-time or near-real-time data will be fetched and processed to ensure the most up-to-date information is available to the users. The application will also include data visualization libraries and tools to create visually appealing and interactive charts and graphs.


The overarching goal of this project is consolidating market data into one central hub and draw insights from the data. Using this consolidated data, we wish to:  

Centralized Visualization Tool

Build a centralized visualization tool to understand the trends in sentiment from various media houses and gather insights from natural language processing models. 

Use of LLMs to Generate Stock Analysis

To feed current trends and up to date financial news to a Large Language Model (LLM) with the use of LLama models to generate stock-specific analysis in plain English, bridging the gap for new investors.

Visualization Algorithmic Trading Strategies

Visualize algorithmic trading scripts and understand their performance on past data. In addition to past performance, use such scripts to visualize future returns. 

Navigating the Markets

Copyright © 2024 HKU-CS