Week #5 #

Feedbacks #

• Feedback collection plan

1. Target Audience:

   • A-Level economics and business exam students
   • Tutors preparing students for A-Level exams

2. Methods:

   • Surveys: Distributed online surveys with specific questions focusing on user experience, feature utility, and areas for improvement.
   • One-on-One Meetings: Conducted in-depth meetings to gather qualitative feedback and deeper insights.

3. Areas of Focus:

   • User interface and experience
   • Content relevance and comprehensiveness
   • Ease of navigation
   • Performance and speed of the website
   • Overall satisfaction

4. Algorithmic Feedback Collection:

   • Developing online forms to collect standardized feedback.
   • Implementing online charts for quantitative assessments in future iterations, once a larger sample size is achieved.

Next Steps:

1. Expand the feedback network systematically.
2. Continue the feedback - reflection - improvement cycle to enhance the product iteratively.

• Conducted user surveys or feedback sessions

Feedback Collection Activities:

1. Video Demonstration:

   • Presented a comprehensive video showcasing the website’s functionality.
   • Ensured users understood the features and potential benefits.

2. Local Testing:

   • Provided Docker access for users to test the website locally.
   • Enabled hands-on experience to gather practical feedback.

3. Meetings:

   • Conducted online meetings with users to discuss their experiences and gather feedback.
   • Allowed for real-time interactions and immediate clarification of any issues.

4. Surveys:

   • Distributed a concise survey to collect structured feedback.
   • Focused on key aspects such as usability, functionality, and overall satisfaction.

Key Insights:

• Real-time feedback during meetings provided in-depth insights into user experiences.
• Local testing via Docker allowed users to explore the website’s features comprehensively.
• Survey responses supplemented our understanding of user satisfaction and areas for improvement.

Next Steps:

1. Analyze the collected feedback to identify common themes and actionable items.
2. Implement necessary improvements based on user input.
3. Continue iterative feedback sessions to ensure ongoing enhancement of the product.

• Analyzing feedback, identifying and prioritizing issues

1. Documentation:

   • Compiled all feedback from surveys and meetings.
   • Ensured clear identification of areas for improvement or enhancement.

2. Analysis:

   • Reviewed feedback to identify common themes and patterns.
   • Highlighted recurring issues and user suggestions.

3. Prioritization:

   • Assessed the impact and feasibility of implementing the feedback.
   • Focused on high-impact, low-effort improvements first.

4. Team Meeting:

   • Scheduled an internal meeting to discuss feedback findings.
   • Evaluated the potential benefits and drawbacks of proposed features.
   • Made decisions based on resource availability and project timelines.

Key Insights:

• Users are generally satisfied with the current functionality of the product, especially marking the topic classification feature and generating the new exam feature, which is especially useful for the A level exam tutors.
• Identified minor enhancements that could improve user experience without significant resource investment.
• No major new features will be incorporated at this stage due to user satisfaction and resource considerations.

Next Steps:

• Implement feasible enhancements as identified during the feedback analysis.
• Continue monitoring user satisfaction and gather additional feedback in future iterations.
• Re-evaluate the need for new features as the project progresses and based on evolving user needs.

Roadmap: #

Current Status:

• Team undecided on post-Capstone involvement and is currently inclined not to continue with the project.
• No extensive roadmap established yet.

Potential Future Steps:

• Decision Point: The team will discuss and decide on continuing the project towards the end of the semester.
• Roadmap Development: If we proceed, we will create a comprehensive roadmap including short-term and long-term goals based on feedback.

Conclusion: While no immediate roadmap is being created, the team is open to future development. A structured plan will be formulated if the project continues beyond the Capstone phase.

Weekly Progress Report #

Backend #

• Continued unit and user testing.
• Completed extract parser development.
• Finalized and tested exam generation feature.
• Implemented constant integration logic.
• Connected new front-end pages to backend functionality.

Frontend #

• Finished converting the Figma prototypes to code for creating, editing, and uploading questions.
• Connected all pages to backend.
• Enhanced code formatting for readability.
• Developing the home page, single question review page, and mobile adaptability.

Machine Learning #

• Conducted model testing to prevent data leakage and ensure accuracy.
• Progressing on the interpretability of model decisions using Kolmogorov-Arnold Network, which requires extensive training time.

This week, we made significant strides in backend, frontend, and machine learning development, aligning our efforts towards a cohesive and efficient product.

Challenges & Solutions #

Challenge: Significant training time for Kolmogorov-Arnold Networks (KANs).

• Solution: Allocated dedicated resources and optimized training schedules to manage the extensive training period.

Challenge: Identifying the correct logic for CI/CD pipelines.

• Solution: Scheduled focused sessions for the coming week to resolve CI/CD pipeline issues and ensure proper implementation.

Initial Demo #

Below is the initial demonstration of the functionality of the product. All the pages are working and had been connected to the back-end.

Conclusions & Next Steps #

Conclusions: This week, we made substantial progress in backend development, frontend integration, and machine learning model testing. Despite facing challenges with the training time for Kolmogorov-Arnold Networks and CI/CD pipeline logic, we have outlined clear solutions to address these issues.

Next Steps:

• Resolve CI/CD pipeline logic issues.
• Continue developing and refining the home page, single question review page, and mobile adaptability for the frontend.
• Proceed with the interpretability development of the Kolmogorov-Arnold Network.
• Maintain ongoing unit and user testing to ensure product stability and performance.