Mouse Dynamics
Overview
Mouse dynamics analyses how a person moves and clicks a pointer. Movement speed, path shape, and click timing all vary between individuals and can form a behavioural fingerprint.
What is measured
Students click through 8 on-screen targets in sequence. For each segment between targets, three features are captured:
| Feature | Description |
|---|---|
| Movement time | Time taken to move from one target to the next, in milliseconds |
| Path curvature | How curved the movement path is (ratio of arc length to straight-line distance) |
| Click dwell | How long the mouse button is held down on each target, in milliseconds |
Enrolment
Students complete the 8-target sequence the required number of times (default: 5). Each run produces a vector of movement times, curvatures, and click dwells. The profile stores the mean and standard deviation of each feature across all attempts.
Identification
The system computes a normalised Manhattan distance between the sample and each stored profile, then converts distances to confidence percentages using softmax with a configurable scale factor. See How It Works for the formula.
Configuration
All settings are managed via the Admin Panel — no code changes required.
| Setting | Default | Description |
|---|---|---|
| Enrolment attempts required | 5 |
Number of target-clicking runs required to build a profile |
| Confidence sensitivity | 2.0 |
Softmax scale factor for converting distances to confidence percentages. Higher values sharpen the separation between matches. |
Discussion points
- Try the module with a mouse vs. a trackpad — the profiles are noticeably different.
- How would the system perform if a user was tired or in a hurry?
- Could someone learn to mimic another person's mouse movement?