A Word-Predict App

Jiandong Jin
2015-04

Introduction

I've developed a word-predict app that takes as input a phrase (multiple words) in a text box input and outputs a prediction of the next word. Check this app here.

A N-gram model was trainned with the datasets given in the course project. N-gram models can be imagined as placing a small window over a sentence or a text, in which only n words are visible at the same time. The simplest n-gram model is therefore a so-called unigram model. N-grams start to become useful when n is two (a bigram) or greater.

In this App we using 5-gram model to predict.

How to use this App?

  • Simply enter your text into input panel
  • Right-sidebar will show the text you've just entered and predicted next-word.
  • It can predict dynamically when you keep entering text.

Pipeline

Future Work

  • This App is lack of analysis of performance, such as time spend and accuracy, I will still work on it and update my App.

  • I should use more plots to show my predict results, and use dynamic chart to visualize more predicted word sorted by probability.

  • Shinyapp UI's design is a hard work for me, I have to learn from many others shinyapp, I still need more pactice to improve my HTML/CSS and designing skill to make my report more attractive.