Predictive Experiment - Mini Twitter sentiment analysis
This experiment demonstrates the use of the Execute R Script, Feature Selection, Feature Hashing modules to train a text sentiment classification engine.
This specific experiment is a second step in creating a web service for [twitter sentiment analysis](http://gallery.azureml.net/Details/397277b61137467086a18eb0e3d47e34). In this step we will convert training experiment to predictive experiment.
The steps to accomplish this:
1. Go to your [training experiment](http://gallery.azureml.net/Details/397277b61137467086a18eb0e3d47e34) and click create predictive experiment under bottom menu "setup web service"
2. Go to your predictive experiment (that is this experiment)
3. Add project columns module to remove sentiment label column
4. Modify execute R experiment to skip using sentiment label
5. Make sure feature hashing bits and n-gram matches the one in training experiment
6. Run the experiment
7. Publish experiment as web service by clicking deploy web service on bottom of the screen
Once above steps are completed, your graph will look like the one shown below
