COMP4801/FITE4801. Final Year Project 2022-23
Undergraduate course, Department of Computer Science, The University of Hong Kong, 2022
Student individuals or groups, during the final year of their studies, undertake full end-to-end development of a substantial project, taking it from initial concept through to final delivery. Topics range from applied software development to assignments on basic research. In case of a team project, significant contribution is required from each member and students are assessed individually, such that each student is given a separate project title. Strict standards of quality will be enforced throughout the project development.
Spot A Doc
In Hong Kong, scheduling an appointment with a doctor can be challenging due to high demand and shortage of medical professionals, leading to long waiting times on phone calls. This not only results in a negative user experience but also significant time wastage. Moreover, many clinics lack an online presence or centralized platform to provide appointments or share clinic and doctor information with the public. Consequently, people find it challenging to locate a suitable healthcare provider that matches their needs. Our project aims to provide a solution to these shortcomings by introducing an e-commerce platform for commercial clinics, named by the project team as Spot A Doc, to facilitate convenient online appointment reservation, as well as cryptographically verifiable, digitally signed medical certificates, in the commercial realm.
NFT Fraud Detection System
An industry-based project to build a NFT Fraud Detection System in collaboration with DTTD (Dotted), a local start-up developing a mobile-first NFT social platform. The popularity of NFTs has drastically changed how people transact assets, but it also gave rise to many artwork scams and thefts, which has undermined the benefits of creators and investors, as well as the stability of the NFT ecosystem. Given the problems caused by fraudulent NFTs, this project aims to build a NFT scam detection system and find the patterns or relevant scams. The potential outcomes are to identify fraudulent tokens and prevent unauthorized activities.
The Future of Non-Fungible Tokens in the FinTech Space: Potential Applications and Trends
In this ever-changing era, the FinTech industry is continually shaping the whole world, especially many sectors of the finance industry. New concepts and their applications are springing up. One of the most lately popular topics is Non-fungible Tokens (NFTs). Many people see them as very fancy digital designs and beautification. The nature of NFTs is a digital security storing particular data (e.g., ownership) based on blockchain. As a new digital asset, NFTs have received intense attention after their exposure to social media because of the social influence of celebrities and their crazy sell amount. Having various application scenarios like art collectibles, intellectual property, patents, the metaverse world, the gaming industry, the supply chain, etc. (an image of NFT artwork is shown below, according to WSJ), NFTs are not a new thing, and they are the products of more than one decade of development, iteration, and progress of blockchain technology and digital currency. In this final year project, which may not have too much programming view but track the cutting-edge opportunity trend of FinTech application away from school courses. The main focus will be a deep-down reflection on the NFT ecology and its development potential to face opportunities and challenges from the future. In this project, the author will try to find out what the profound logic behind NFTs is, which areas and scenarios will a flash in the pan after NFTs’ current hype and bubble be, which great strides and ultimately change the world may make, and eventually, what possible ideal form of NFTs could show.
Building a mock NFT marketplace for learning AI and Blockchain technologies
Non-fungible tokens, NFTs are digital assets built on blockchain technology that have constantly been booming since early 2021, with average monthly sales of approximately 20 thousand. However, despite its hype and popularity, the general public implicitly lacks an understanding of the principle and structure of the cryptocurrency ecosystem, product, and technology itself. Learning in those fields will be very important as blockchain technology is often praised as the internet’s and data science’s future. Therefore, this project proposes adopting two of the most evolving technologies to build a web application that provides end-to-end user learning experiences through a mock NFT marketplace. For the first part, when a user enters a description of the image he/she wants to create on the platform, Artificial Intelligence technology will automatically generate an image that matches the user input. We will use an open-source AI called Dall-E; we plan to customize and apply it to suit our project’s purpose(AI, NFT education). While the AI creates images, users can learn the working principle of AI with educational materials collected from different sources through our platform. With the image generated, the user can mint an NFT, where blockchain technology comes in. In this process, all the image’s information is pushed to the blockchain, thus creating the NFT, a unique address storing the data. We will build a mock marketplace on the Klaytn development network on which users can get coins without any real monetary value to buy and sell other users’ AI-generated NFTs. One of our top goals is to educate on NFTs technology; we will craft an intuitive frontend providing step-by-step guidance and explanations and giving a high-level overview of the operations happening on the blockchain.
Web3 Health Insurance
By 2028, the health insurance market, which was valued at 1.9 trillion US dollars in 2021, is projected to rise at a 7% annual rate.The issue, however, is that insurance firms do not factor in users’ healthy lifestyles when calculating monthly costs. The failure of the medical and health insurance sector to make use of the enormous amounts of data at their disposal is yet another problem. Medical institutions, health insurance providers, and other players in the greater health sector are not digitally native; they frequently still rely on antiquated bureaucratic procedures, and digitalisation has not resulted in an efficient revolution for them. The goal of this project is to build a blockchain-based mobile application that tackles these two issues. In order to do so, the application is characterised by two main features: an activity based crypto reward system and the the feasibility for users to sell health data to interested parties who can obtain and validate this information via Arkam. These features are implemented through a combination of industry-standard technologies and frameworks on the server-side and the client-side. Background research on similar projects in the domain was conducted. The primary takeaway is that these applications’ technical designs are poor, they offer ineffective incentives, and they simultaneously use flawed tokenomics. Arkam addresses all of these issues while providing the users a full fledged user experience with insurance and complete control of their data. The system’s main building block is a set of smart contracts driven by Solidity. The program has all the features one would anticipate from a regular insurance provider, plus much more. The ultimate objective of this application is to make the wealth of health data conveniently and securely available while also ensuring that users aren’t overcharged for insurance despite leading an active and healthy lifestyle.
Predicting Interest Rate with Machine Learning
Interest rates play an important role in the economy and financial markets. In the economy, interest rates affect the willingness of household consumers and business institutions to spend or save, which in turn affects the economic output and economic growth of the country. Interest rates in an economy also affect the willingness of investors to invest using a specific currency relative to others, which has an impact on exchange rates, affecting the profitability of firms that has exports or imports from other countries. In financial markets, interest rates are used in many financial models (such as the CAPM model) to model prices for stocks and other financial products. As a result, there are many parties that are interested in knowing the future direction of interest rates so that they can make informed decisions. This project aims to build a machine learning model that predicts the change in the 3-month US treasury bill rate based on features extracted from FOMC meeting minutes. Different methods of extracting features from the text data and different machine learning models will be tested and compared.
Blockchain NFT project
The project aims to build an advanced system for the voluntary sector in Hong Kong. It aims to provide “show off” value for volunteers. And to provide managing value and enhance service quality for organizers. It will be blockchain-based together with a physical verification tool. That will give out NFT when someone does volunteer work. Also, track and store one’s performance when doing volunteer work. Therefore a web application will be built for organizers to view and manage volunteer work. The website should also with good UI/UX for smartphone users.