Text prediction shiny application created for
Coursera-Johns Hopkins University
Data Science Specialization
Switftkey Capstone Project
Author: Alejandro Mut
Date: Sep 17th, 2020
The aim of this application is to meet the need of a text predictive Shiny application which can predict the users next word with speed and accuracy with the emphasise on the latter.
“It will not matter how fast the app is, if it is inaccurate.”
The algorithm is based on a N-Gram model that was built from a large corpora supplied by SwiftKey. The material was sourced from Twitter, News and Blogs.
| Total Size | Total Words |
|---|---|
| 28MB | 4,500,000 |
Woracle uses a clean minimal user interface (UI) which access the data compression and predict algorithems working in the background.
The user enter text into the input box and waits for the next word to be predicted.
Fast and simple, the algorithm is build for accurancy first, then speed. The average typist types 36WPM, the algorithm will work to meet that critiria.
accurancy: 60%
speed: 1 to 2 secs
memory used: 4MB
hard disk used: 50MB