John Hopkins - Coursera Capstone Pitch-Slides

The goal of this exercise is to create a product to highlight the prediction algorithm.

• Algorithm for predicting the next word given one or more words using Natural Language Processing.
• The data used for this project was provided by swiftkey in association with coursera. A large corpus of blog, news and twitter data was loaded, cleaned and analysed.
• N-Grams up to 5 were created and used with back off techniques to build this project

I am using the NLP technique with stupid back-off strategy to build this predictive model.

The quanteda and tm package are used extensively.The quanteda pakcage is more useful in creating N-Grams as they use less memory.

The tm package was used for building the corpus and cleaning the corpus of stopwords, puntuation, number and special characters etc.

1 to 5 - Grams were built,stored and used with a backoff strategy to predict the next word

-Provides a input text for the user to input a word or a sentence -The app will start predicting as soon as the something is entered in input text box. =The best possible next word for the input text will be displayed. -Work is in progress to provide best 5 words based on the input text you haven't tried out the app, go here to try it!

• Predicts next word as the user types a sentence
• Similar to the way most smart phone keyboards are implemented today using the technology of Swiftkey

The quanteda and tm package are used extensively.The quanteda pakcage is more useful in creating N-Grams as they use less memory.

The tm package was used for building the corpus and cleaning the corpus of stopwords, puntuation, number and special characters etc.

-http://www.cs.columbia.edu/~smaskey/CS6998-0412/supportmaterial/langmodel_mapreduce.pdf)

• Checks if highest-order (in this case, n=5) n-gram has been seen. If not “degrades” to a lower-order model (n=3, 2); we would use even higher orders, but ShinyApps caps app size at 100mb

-The app can be accessed form here: https://stevedubois.shinyapps.io/app50/

-The presentation/pitch/slides is here: http://rpubs.com/steveduboi/378308

-The code for the project can be accessed from here https://github.com/steveduboi/Capstone_Project.git

-Other Information re: NLP, Katz-Back-Off info: https://en.wikipedia.org/wiki/Katz%27s_back-off_model http://www.hlt.utdallas.edu/~sanda/courses/NLP/Lecture06.pdf