Data Science Specialization : Capstone Project

This is a NLP problem focused in text mining. This project allows ground work on raw data for text prediction.

About :

• Text mining project to predict next word
• Uses R programming
• App made by Shiny and hosted at shinyapps.io

The model is based on Tidy Data principles in R. The steps involved are :

• Input: text files as training data
• Clean data: Group data into 2, 3, 4 word n-grams
• N-grams function: uses “Katz back-off” prediction model
  • last 3, 2, or 1 words to predict the best 4th, 3rd, or 2nd match in the repos
• Output: Next word prediction

Provides a simple UI to the word prediction model.

Salient Features:

• Predicted outputs dynamically below user input
• Tabs with plots of most frequent n grams in the data-set
• Allows large training data for future enhancement
• Fast output response

• Hadley's Tidy Data “http://vita.had.co.nz/papers/tidy-data.html”
• Text Mining with R: A Tidy Approach “http://tidytextmining.com/index.html”