The objective is to determine the emotion of a person taking a survey by tracking mouse and keyboard. For example, if the response time between opening the page and answering is long, the person is not confident. The data collected will be fed to Machine Learning algorithms to visualize the data and recognize emotions from them.
Run interview.py and open the page "/admin" (for admins) it at the local host with port connected.
To begin the survey, the admin logs in to the system, adds the candidate name and details. Press record to start recording. Press playback to see the user's recorded survey. A simple admin page is shown:
The admin can then place the candidate on the chair with all sensors (eyetracking, ECG, etc. ) when the browser shows like:
The survey includes questions of different types including:
The mouse- tracking done in the backend gets data in the format:
This survey can be played back in the same motion as the user answered.
interview.py - (main page) opens the admin page and generates questions or playback as per the admin response
admin.py - checks admin ID password
Quiz.py - generates quiz questions based on the quiz ID given by admin
db.py - connects to the MySQL database
dbase.py - functions for accessing and modifications in the database
model.py - functions to create user directory, record and playback
keym.py - functions to record the mouse events
playback.py - fucntions to playback the entire survey
Question- List of all questions
Quiz- quiz ID linking questions
Static- audios/videos/images
Subjects- directory for each subject taking the survey
Templates- HTML templates for the each quiz question