Around the world, people are spending an increasing amount of time on their mobile devices. But typing on mobile devices can be a serious pain. This app pretends to help a smart keyboard to makes it easier for people to type on their mobile devices.

To make the app, the job was to understand and build a predictive text model with: analysis of text data and natural language processing.

This is the Capstone Project in Data Science Specialization in Coursera, develop by The Johns Hopkins University and Swiftkey like the corporate Partner.

Data Preparation
The data used in this project was given by SwiftKey company from blogs, news and twitter. The Fist step was clean the data removing all unimportant words and expessions. After, take a sample to build the model (3% in this case).

The model
Build a Text Mining Model, based in the N-gram technique. Calculating the 'Next most Probability Word' over of frequency of two, three and four combinatios of words in the sample.

Performance
- The data preparation for this model takes more than one hour.
- The sampling reduce 2.5 Millions of lines to +100k.
- The time to find the 'next word' is 0.5 seg in average.

• The app is available in: https://franmarq.shinyapps.io/nextwordpredictor/
  
  Instructions This app was built to better results in English languaje and when you use three or less words.
  
  Follow the next steps:
  1. Write some words in the 'Expression:' box
  2. Press 'Submit' button to obtain the result
  3. The 'next predicted word' will appear on the right side panel

• The app documentation is available in: https://github.com/franmarq/DS-CapstoneProject
  

franmarq@gmail.com