project-proposal-2024

Manage My Assignments

Abstract

One of the most difficult aspects of the University experience is managing multiple assignments and the time put into completing them. Assignments which initially seem simple can end up taking entire weeks, implementing basic functionality can be held back by bugs esoteric in origin, and progress on one assignment can be delayed by the need to complete another. The aim of Manage My Assignments is to create a platform which helps students schedule their assignments into blocks of work designed to maximise productivity based on their study schedule, deadlines and assignment difficulty.

Author

Name: Mark Tobys

Student number: 45294417

Functionality

• AI integration: Users provide a description of the assignment overview which is fed into an AI model to predict how long the assignment will take to complete
• Project Breakdown: Provides breakdowns on the assignment timeline including estimated time required for planning, functionality implementation and bug-testing
• Scheduling System: Creates a work schedule for a given assignment based around a students current work/study schedule, assignment deadline and other assignments
• Troubleshooting: provides insights on common bugs which may occur during development, and resources to help with bug and error testing.

  Scope

  In terms of a minimum viable product, Manage My Assignments should be able to do the following:

• Accept an assignment description and use an ML model to generate a timeline breakdown for different stages of the assignment
• Interface with the UQ timetable API to obtain a students study plan and generate a work timetable for the assignment based around classes/the assignment due date
• Use an ML model to provide resources to help with each block of work (planning/development/bug-testing)

Quality Attributes

• Availability: Manage My Assignments should be available to access across a variety of higher education institutions across Australia and potentially globally
• Interoperability: Manage My Assignments should be able to interface with University scheduling applications and ML models to obtain student timetables and generate timeline breakdowns for assignments
• Reliability: The estimated completion times provided by the ML model should be consistent, i.e. if 100 students ask for timeline breakdowns for the same assignment it should produce the same answer every time
• Scalability: The application should be able to handle surges in requests without breaking, for example if a class assignment is released, all students enrolled in the course should be able to request an assignment breakdown at the same time without the system crashing

Evaluation

• UQ and ML API interface: The Development team should ensure that the application can reliably communicate with the UQ/UQ Timetable and ML model API to send and receive information and that the application is able to successfully hand errors
• AI interpretation of assignments: The development team should use assignments they are intimately familiar with to evaluate assignment breakdown timelines and schedules to determine the effectiveness and accuracy of the ML model in estimating the required completion time for assignments
• AI consistency and reliability: The development team should thoroughly test the ML provides consistent estimations for assignment breakdowns to ensure that the same results are provided to all students undertaking the assignment, and that the ML model can produce consistent, correctly formatted feedback which can be extracted by the application API for interpretation
• Scalability Testing: The development team should test the application under heavy load to simulate situations such as a class of students requesting an assignment timetable breakdown upon the release of a new assignment
• Timetable Implementation: The development team should test the scheduling system with multiple overlapping assignments to ensure that the scheduling system can effectively allocate work periods for each assignment without conflicts/clashes across reasonable time frames