Honesta
risk-based proctoring · maximum marks: 50 (lock)
- Candidate:
- Vishwas K
- Duration:
- 2025
- Centre:
- Solo
- Result:
- Finished
Q1.What does this project do? (Choose one.)
[1]Q2.Why is a project like this needed?
[2]Online exams are easy to cheat on, and manual invigilation doesn't scale. An online exam platform with real-time proctoring: the browser streams student behaviour over WebSockets, the server keeps a running risk score, warns the student and locks the exam past a threshold — and examiners can generate MCQ papers with Gemini.
+2 goodQ3.Describe how the system works, step by step.
[4](a) Sign in — Local or Google login with Postgres-backed sessions.
+1(b) Observe — The exam page listens for tab switches, keystrokes, mouse movement, copy-paste and screenshot keys.
+1(c) Score — Events stream over Socket.io and add weighted points to a server-side risk score.
+1(d) Enforce — Past 20 the student is warned; past 50 the exam locks permanently.
+1
Q4.Write short notes on its features.
[8]Server-side risk score
Weighted events with cooldowns, decided on the server.
Warn then lock
Two thresholds, enforced in real time.
Copy & screenshot guard
Copy-paste blocked and the clipboard wiped on PrintScreen.
AI question papers
Gemini generates validated four-option MCQs.
Q5.Show your working. Where in the code does this happen?
[4]Server-authoritative scoring: Weights, cooldowns and thresholds live on the server, not the client.
ref: backend/server.js ↗Client proctoring hooks: Visibility, input and clipboard listeners wired into the exam page.
ref: frontend/src/components/ExamPage.js ↗Materials permitted
Frontend: React 19, Bootstrap, Recharts, Zustand · Backend: Node · Express, Socket.io, Passport · Data & AI: PostgreSQL, Gemini 2.0 Flash