Capstone Project - Slide deck presentation

Nordine RAJAOUI
08/11/2020

The Project

The goal of this project is to build a natural language processing model that will predict, given a sequence of words, the next word of the sentence.

Project deliverables:

Prediction Model

Given the three files, we will create the bigrams, trigrams and quadgrams. Then, we will save those ngrams on rds files to be used by our application.

The application will:

  • Clean the input
  • Count the words in the input
  • If input contains 1 word, search last word of bigram starting by the input.
  • Else if word contains 2 words, search last word of trigram starting by the input.
  • Else search last word of quadgram starting by three last words of the input.

Next word application

The application contains an input box with a submit button in the left frame. The right frame contains a tab that will display the predicted word.

The button in the left frame has been put to avoid confusion with last submission (the answer takes a few seconds).