Homework Wizard / 2024
Overview
This sponsored project, a collaboration between SAS and NC State, centers on visualizing complex systems. Homework Wizard is a digital platform designed for educators to monitor student progress, pinpoint learning gaps, and assign personalized homework. By integrating APIs with the school’s existing curriculum and textbook databases, the platform leverages a Large Language Model to generate custom questions, provide insights into student responses, and adapt to content-based learning patterns.
Using SAS App Factory as a probable workspace, select capabilities of an education app and connect to them using APIs while creating a UI built for end users to access. Understand the data structures needed to inform the selected industry (education) and correctly identify functions needed for the proper creation of the interface.
Project Objectives
I designed the student view for Homework Wizard, a digital platform aimed at helping educators track student progress, identify learning gaps, and assign personalized homework. I also conducted extensive research on machine learning and its implementation within the system, utilizing APIs to integrate with existing curriculum and textbook databases.
My Role
Research
The research process included developing user personas to understand the needs of both educators and students, as well as interviewing an elementary school teacher to gain deeper insights into real classroom challenges. I also conducted extensive research into AI and machine learning to explore their potential in education. This technology would be integrated using APIs to connect to the school’s existing curriculum and textbook databases, generate custom questions, offer insights into student responses, and adapt to content-based learning patterns.
Within the education sphere, our main demographic for this platform is elementary school teachers and students. Both users have an individual platform that is connected within the database to form a teacher-student relationship.
Users
Mrs. Anderson is a dedicated 3rd-grade teacher who strives to pinpoint learning gaps and streamline her workload while effectively managing a classroom of 26 students.
Persona
Better identify learning gaps and monitor student progress in her classroom.
Lessen her workload throughout the day.
Be able to accommodate to each student’s needs.
Evaluate the efficacy of her lesson plans over time.
Goals
A majority of our research revolved around how we could theoretically implement AI, APIs, and functions in our program. Not only did we make a UI that functions on the frontend, we also planned out how the backend would work as well.
Extracted data from student profiles can be used to quickly create data visualizations which categorize grade averages, trends, and outliers. Educators also have the ability to connect their grade book and automatically record homework grades, as well as utilize data from students' other assignments to visualize cumulative grade trends.
Backend Planning
Generative AI is utilized to write questions tailored to each student’s needs. Student profiles consist of their previous answers, average response time, grades, weakest subjects, and preferred learning styles, and are stored in a database within App Factory.
Machine Learning would be used to analyze student profiles for their strong and weak points, and students will be given more relevant questions. For example, if a student is doing well in English but poor in math, they will be assigned more math problems to practice.
Design Process
The most challenging part of this project was figuring out how to organize student data into a simple UI. It was easy to overcomplicate the visuals, but eventually the use of buttons to expand information made the UI less cluttered.
Mrs. Anderson logs on to Homework Wizard and quickly sees how well students completed their homework the night before. After seeing entire class averages, she uses the “Magic Wand Tool” for in-depth insights and spots outlier profiles that need attention.
Taskflow
The most challenging part of this project was figuring out how to organize student data into a simple UI. It was easy to overcomplicate the visuals, but eventually the use of buttons to expand information made the UI less cluttered.
Wireframes
