Big Data Review Analysis

Can tell you the difference between positive and negative movie reviews using Machine Learning.

Keep in mind that it took me several days/weeks and beers to create this project So please be kind and give me credits for this code. I won't bite

Ready, Set,.. LEARN!

Now that you know your legal rights after reading the 2 lines above we can finally start this journey to machine learning and letting your computer tell you if a movie review is positive or negative.

But before we continue, allow me to introduce myself. My name is Naiba.

Naiba is very nice and willing to learn, but you need to train her. It's just like Pokemon. Except that she works Naive Base and the you can train hey from data in datasets instead of TM's of your TM CASE

In order to use Naiba, simply run main.R. Now you can call a variaty of commands, let's start at the beginning:

Build a skillset

• If Naiba already learned something, call learn.load()
• If Naiba doesn't know anything, call: learn.teach()
• To make the program even smarter (Append skillzz), call: learn.load() first and then learn.teach()

Test our skillset

1. Call the sentiment.test() function.
2. Dare Naiba to make the tests in your set
3. Tell what to make
4. Await results
5. If you run this on a Mac you can use Threading by running the threaded.R file and calling: sentiment.test.threaded() (This will currently not show a progressbar but is much faster, duhhh)

Test your own review

1. Call the sentiment.calc(<your review string>) function
2. After sometime, Naiba will tell you if the review is positive or negative

Save what you learned

• Call the learn.save() function
• Now Naiba will remember whatever you learned her. So she can go to sleep safely.

That all sound nice, but what are all the functions so I can mess with them myself?

You just need to ask nicely:

• console.ask(string, type=string) => Internal use only, ask for a string or integer
• console.confirm(string) => Internal use only, ask a true/false question
• learn.load() => Let's dig in her memory to see what she learned before
• learn.save() => Let her remember everything you just learned
• learn.teach() => Teach her new stuff
• sentiment.calc(string, progress=boolean) => Check if a string is positive or negative, with or without a progressbar
• sentiment.split(string) => Internal use only, splits a string
• sentiment.test() => Dare to test her
• sentiment.train(string, 0|1) => Learn her that string is positive (1) or negative (0)
• set.import(string) => Import data from given path

Threaded functions

Currently these only work on the best computer system there is: MacOS Naiba can be so much faster when she uses all the cores of her brain (CPU). Call these and set her free

• sentiment.test.threaded() => Dare to test her. Progressbar currently NOT showing