NextWord

Chilly Amador
01/24/2016

What is NextWord

  • NextWord is an application to predict the next word the user will type based on the previously typed words
  • It was developed as part of the JHU-Coursera Data Science Specialization
  • It used large text samples of Twitter, Blogs and News websites to create its prediction algorithm

alt text

How NextWord Prediction Algorithm works

  • NextWord algorithm is based in n-grams detection and matching
  • An n-gram is a contiguous sequence of n items from a given sequence of text or speech
  • NextWord ngrams are word sequences of 2, 3, 4 words
  • NextWord uses a n-gram corpus extracted from the internet (Twitter, Blogs and News) to identify the next word that will be typed by a user
  • NextWord normalizes the text to lower case, eliminates punctuation, whitespaces, numbers, URLs and stopwords (determinants, conjunctions, prepositions)

How to use NextWord

  • The NextWord App is easy to use
  • Type a message in the text box
  • Click on the bar that says: “Click here for the next word…”
  • Below the bar, NextWord will display all the words that were predicted as the next word to be typed
  • To restart, erase the message/text and type a new one.

Try NextWord

chillyamador.shinyapps.io/CAPSTONE_PROJECT/

alt text