Data science spesialization Capstone project

Application was build for end Data science specialization of john Hopcins university on Coursera

link to application:

https://nikkrichko.shinyapps.io/Prediction_jhu/.

link to preparation description:

http://rpubs.com/nik_krichko/297116.

The biggest part of application was in preparation of Ngram. And typical way of preparation are:

• read data from file
• get random 10 persent of all values
• clean sample from trash

• create Ngram's you can read more how it was on link

  http://rpubs.com/nik_krichko/297116.

  And the most of work for prediction was do library - ANLP

  https://cran.r-project.org/web/packages/ANLP/index.html.

  it was necessary only add to function predict_Backoff input text and list of prepared Ngram

contain:

• input text field

– just neccessary add new words and wait

• lower (inactive) buttons with predicted words.

– works without stop. Algorithm add all possible variants on that buttons with some delay.

contain some function:

1 At first it read prepared gnrams

2 render buttons with refault text

3 Wait unlit user type something in text field

4 start predict using (library(ANLP))

5 and change button labels

thats all