Implementation and evaluation code for Key-HIT

Paper title: Key-HIT: Hash-Induced Transformation for Keystroke Dynamics-Based Private Continuous Authentication
Authors: Yatish Dubasi and Qinghua Li

Requirements:
Ensure that you install any Python modules used within our Python scripts

Code descriptions:

• IKDD: the original dataset from https://github.com/MachineLearningVisionRG/IKDD
• IKDD_split: dataset with training and testing split
• basic_testing.py: tests the typing_operator.py functions and displays distances between 2 users
• comparative_tester.py: evaluation script that presents timing and accuracy results for various methods
• data_counter.py: simple script to count how many users we can utilize given desired training and testing counts
• decoded_features.txt: individual keycodes and digraphs used as keystroke features (human-readable)
• features.txt: virtual keycodes and digraphs used as keystroke features
• parse_data.py: extracts split dataset given desired training and testing counts, along with general dataset info
• testing_helper.py: used with comparative_tester.py to process enrollment and authentication
• typing_operator.py: Key-HIT methodology within a python class that can be imported in other scripts