-
-
Teacher View
-
Student View
-
Login
-
Sign Up
What it does
Eproctor is an application designed to prevent verbal abuse over the Internet. Due to the COVID-19 pandemic, a high number of students have moved to an online education. While virtual classrooms have given teachers an opportunity to interact with students remotely, a lack of physical supervision allows students to misbehave. Eproctor helps by monitoring students' conversations between peers and between teachers, then analyzing for abusive language. A built-in warning system alerts the student and and the teacher can check what the student has said at any time. Some people may argue that this is an invasion of privacy, however, Eproctor can be seen as an application that runs alongside other programs, such as Zoom, that already record students under school policy. In the end, Eproctor can be seen simply as another layer of security to ensure the wellbeing of all students and teachers.
FAQ
Are there privacy issues with recording? Students can be required by the teacher to start recording at the beginning of Zoom meetings. By looking at timestamps, the teacher can determine that the student is in the session. If the student forgets to close out of Eproctor at the end of the session, data is securely protected by Firebase authentication and the teacher can remind the student.
What if Eproctor incorrectly flags offensive language? At the end of the day, the teacher has the jurisdiction over issuing discipline to the student. If the teacher determines based upon context of the recorded offensive language that it isn't offensive, the teacher doesn't have to issue punishment. Context detection is a future feature that we wish to implement into Eproctor.
Doesn't this require a large amount of processing? The Python GUI and NLP algorithms are very lightweight and can run with minimal processor utilization.
How we built it
We used Python for the entire project. We researched about various modules on python that help Eproctor give its distinct features, and found out that Python works best for, given our very limited limited time. We are using a module that utilizes the google translate API to recognize speech, and a module from GitHub to look for abusive language.
Challenges we ran into
We wanted to use react js to make this project, but within 2 hours of starting, we realized we are not proficient enough to rapidly prototype with react js. For this reason, we PyQt5, a python based UI module for our interface. we also had trouble with tying all of it together.
Accomplishments that we're proud of
We are still surprised by how much we got done, and we strongly believe this project has potential in the real world. Cyberbullying is a raising issue, and we developed this app to counter that. although it may not be a full-fledged application, it is a start, and we are proud of what we have made.
What we learned
We learned a lot about developing python applications. we also learned about more advanced concepts and made use of them, such as threading.
What's next for eProctor
We plan to develop this into a full app, making it work alongside online conference applications like Zoom, google meet, and Teams. we also want to make this into a web application, so that it will not take up too many resources on the user's computer.
Log in or sign up for Devpost to join the conversation.