Assignment dumps in a limited span are quite taxing on students
The less transparency between subjects and untracked progress are the main reasons for this issue, leading to workflow conflicts and a hit on efficiency.
Asgn. aims to solve both of these problems in the following ways:
- Increasing cross-subject transparency by introducing a global calendar for teachers with ML generated deadline suggestions based on students data sets and other subject workloads.
- Tracking and improving student progess by model-generated assessments that are purely for improving core knowledge of the subject.
In addition to this:
Asgn. consists of both a student and teacher side
The teacher side
- Teachers may create and upload marked assignments
- Teachers may have a look at the global calendar
- A home feed to see latest posted assessments along with ML generated ones.
The student side
- A student home feed to see latest posted assessments with a due date and estimated completion time
- Deadline notifications
- Distinct colour coded subject calendar
Two most significant challenges we encountered were:
- Moderate difficulties making our app look great on all devices and screen sizes and responsive.
- Dataset limitations when it came to parameters and algorithm constraints.
- Due to limited data and time constraints, application of KNN model was not possible, leading to lesser accuracy.
- Implementing a complete interface within a short time span