Unreliable data
Manual logs depend on memory and motivation. Students forget entries, round them up, or stop logging when they fall behind.
Sessions are recorded automatically, with idle time removed.
A student project, for student life
Find your study rhythm.
Not someone else's.
Some days you find your flow. Other days, getting started is enough. This research explores a quieter way to understand your study habits, without manual logs or comparing yourself to anyone else.
Opens the EthosTrack app in a new tab.
A week of active studysample pattern
4metrics
0manual logs
1personal score
Abstract
University students often struggle to study consistently. Existing tools depend on manual logging, which is unreliable, or collect data without clear consent.
This system autonomously records login, logout and active duration, excluding idle periods. From that record it calculates a transparent, non-comparative Study Consistency Score built from four metrics: Regularity, Adherence, Stability and Rhythm.
Privacy is part of the design: pseudonymization, k-anonymity generalisation and differential privacy protect the data. The system was validated using an AI-generated synthetic data pipeline.
Three problems this project solves
Most study trackers fail students in the same few ways. Each one has a direct answer in the design.
Uninterpretable feedback
Raw hours and streak counters do not explain whether a study habit is healthy, or what to change about it.
One transparent score, with four metrics you can read and act on.
Weak privacy
Many tools harvest behavioural data without clear consent, and rank students against each other.
Opt-in consent, layered privacy protection and no peer comparison.
Built around everyday study life
Not every study day looks the same. The aim is to help you notice what works for you, not chase a perfect streak. Select a photo for a closer look.
The Study Consistency Score
The score is built from four metrics. It describes your own habit over time and is never used to rank you against other students.
Regularity
How often do you show up to study?
Looks at how many days in a period include real study activity.
- Higher when
- You study on most days of the week.
- Lower when
- Study comes in occasional bursts with long gaps between.
Adherence
Do you follow the plan you set yourself?
Compares your actual active study time with your own target.
- Higher when
- Your active time meets or comes close to your own goal.
- Lower when
- Sessions regularly fall short of what you intended.
Stability
Is your study time steady from day to day?
Measures how much your daily active duration swings around its usual level.
- Higher when
- Daily study time stays in a similar range.
- Lower when
- Long days alternate with very short ones, such as cramming before deadlines.
Rhythm
Do you study at similar times each day?
Looks at whether your sessions start at a predictable time of day.
- Higher when
- Study sits in a familiar time slot, such as most evenings.
- Lower when
- Start times jump between morning, afternoon and late night.
| Metric | Question it answers | What it looks at | Signals a healthy habit |
|---|---|---|---|
| Regularity | How often do I study? | Days with active study across a period | Frequent study days |
| Adherence | Do I meet my own goal? | Active duration against my target | Actual time close to planned time |
| Stability | Is my effort steady? | Variation in daily active duration | Small day-to-day swings |
| Rhythm | Do I study at similar times? | Consistency of session start times | A predictable daily slot |
Privacy by design, in four layers
Each layer removes a little more risk. Select a layer to see how a sample study record changes as it passes through.
Layer 1: Opt-in consentSample record
What happens to your study data?
- You choose.Tracking begins with your consent. You can change your mind.
- Details are protected.Names become identifiers; exact details become broader groups.
- Records stay secure.Protected study records move into secure storage.
- Shared results stay private.Noise is added to shared statistics to help protect individual students.
Tech stack and system architecture
A MERN web application records and scores study activity. A separate offline Python pipeline creates synthetic data to test it.
React.js
Client
The student dashboard: consent screen, session tracking and score display.
Node.js + Express.js
API server
Receives sessions, calculates the four metrics and applies privacy controls.
MongoDB
Database
Stores consented, pseudonymized session records.
Offline Python simulation pipeline
Real student data is not needed to test the system. An AI-generated synthetic data pipeline produces realistic study behaviour for validation.
- GenerateSynthetic student study sessions
- LoadRecords fed into the web system
- ScoreFour metrics and the consistency score computed
- ValidateResults checked against expected behaviour
MMongoDB
ExExpress.js
RReact.js
NNode.js
PyPython (offline simulation)