"Score Intensity of Emotions (Anger, Sadness, Joy, Fear) in tweets"

mediciro

HI all!

我们are currently working on the following challenge:

We need to give an intensity score to a tweet to reflect as much it fits to an label e.g. Anger

Data Structure:

•ID\t TweetText\t Affect\tIntensityScore

•Examples:

•2017-En-22248\t I'm shakingnow.\t fear\t 0.812

The label is given so the target it to predict the score.

Can anyone tell us, how we can set up a score prediction in Rapidminer? We usually need two labels, but we do not know how to create an input, which is giving the modules what it needs.

THANKS

Telcontar120

This is a typical case of text mining with a nominal label (with more than two categories). Take a look at some of the tutorial and training resources for text mining and it should give you some information to get started.

sgenzer

Or you can just use the IBM Tone Analyzer API webservice which will give you all those scores.

https://www.ibm.com/watson/services/tone-analyzer/

Scott