Typing Performance Analyzer

The aim of this project is to calculate typing performance metrics for a given sequence of keyboard interactions. It considers the following metrics:

• Words per minute (WPM)
• Keystrokes per second (KSPS)
• Keystrokes per character (KSPC)
• Error rate (ER)

In addition to these metrics, it distinguishes between following conditions when a user deletes one or more characters and enters another text:

• Correcting a typing error (or typo)
• Writing something else after changing mind

Data Model

The expected input is a sequence of logs of the keyboard events at the character level. In other words, it adopts the transcription sequence paradigm by capturing the entire transcription after every keyboard action. A keyboard log is recorded for every keyboard interaction that either inserts or removes a character. An example log entry is given below:

{
  "_id" : 1619,
  "timestamp" : "1764187044257",
  "device_id" : "4daac81b-fa00-4796-8aaf-6be83de9f9a9",
  "package_name" : "com.whatsapp",
  "before_text" : "This s",
  "current_text" : "[This se]",
  "is_password" : 0
}

Dependencies

• coverage
• nltk
• phunspell

Test & Coverage

coverage  run -m unittest discover -s test
coverage html

Run

python main.py