Touchalytics for Node.js

Touchalytics Node.js implementation.

In my Msc thesis i created a proof-of-concept using the theoretical work from Mark Frank on Touchalytics. Touchalytics objective is to create a user fingerprint based on his usage of a smartphone device. By calculating a set of biometric features using the strokes performed on a mobile device, and inserting it on a classifier, we are able to distinguish from various users.

I divided this implementation in two parts, a client-side and a server-side, this is the server-side in Node.js.

Server.js

Handles the network connections by implementing a Rest API, and controls the program flow.

Configs.js

Contains the various configurations variables.

Extract.js

Performs the calculations of the biometric features using the various strokes received from the user mobile device.

Classifier.js

Utilizes Support-Vector-Machines to train/classify users.

MongoDB.js