Twitter Sentiment Analysis from RapidMiner Template

Ekitsune
Ekitsune New Altair Community Member
edited November 5 in Community Q&A
Dear community,

I am a beginner in machine learning field and I have just started with my first project on sentiment analysis. I have decided to use the template provided by RapidMiner -> 
It works perfectly, however I am unsure whether I have chosen right input data. In the 1st box "Retrieve Historical Sentiment", I put my train dataset (.xlsx) with columns (id, class sentiment, tweet), in the 2nd retrieve box below, "Retrieve", I put my test dataset (.xlsx) with columns (id, tweet). On this test dataset I want to predict the label for all tweets since in the test dataset, class sentiment is missing. 

Results are as following: 




Is this a good approach? Or I am missing something?

I will gladly share with you my project, if needed.

Thank you so much for your help, I really do appreciate it. :)