RSS Feed

Sentiment Analysis made simple

March 27, 2014 by Robin Englebright   

A while back I enrolled on a MOOC module on Data Analytics, and one of the first assignments was to perform a sentiment analysis on twitter. This involved getting a developer API key from Twitter, then writing a python script to suck tweets from the mighty twitter firehose. Analysis involved comparison against a sentiment sorted dictionary of terms.

Whilst those skills are obviously useful, for those not willing to learn Python there’s a cracking tool available to analyse tweets with a hashtag:

http://www.csc.ncsu.edu/faculty/healey/tweet_viz/tweet_app/
I found out about it in a tweet from Randy Olsen, who ran it to visualise the sentiment regarding the recent release of Java 8.

I thought I’d have a look at the #Maharauk14 hashtag and follow  over the next few months as we progress toward the big event. So far there are 5 tweets using that hashtag.

 

Screen Shot 2014-03-26 at 13.51.29

 

The sentiment analysis isn’t very useful on so few, but it’s worth a look at the tag cloud now and to compare it to the final result.

 

Screen Shot 2014-03-26 at 13.51.55

To compare approaches I’ve also set up a hashtag archive on GoogleDocs using Martin Hawksey’s template: http://mashe.hawksey.info/2012/01/twitter-archive-tagsv3/ which does require twitter API keys and what not, but is rather good.

Print Friendly, PDF & Email

»

Leave a Reply

Your email address will not be published.

Skip to toolbar