This workflow examines a sample of tweets from the days surrounding the scottisch referendum for independence in 2014 [1]. After reading the data from a local database, basic text processing is used to extract hashtags from the dataset and term frequencies calculated and used to build a tag cloud. Subsequently, hashtag trending is examined over time, with a notable post election surge in the #the45 movement. Additionally, network analysis is performed in order to look at the most influential social graph surrounding this issue.
Is there any way to use this workflow but not specifically analyze the hashtags for this, but analyze the tweets and make a tag cloud based on that?
I’m not getting anywhere at this point unfortunately.
Are you primarily interested in the tag cloud, or something else? What is the primary analysis you want to do?
This is an older workflow, so you might consider posting a more detailed question on the main Analytics Platform forum, which gets a lot more traffic from our workflow experts
Hey @ScottF ,
Thanks first of all for your reply. We are currently analyzing Twitter data in the context of Data Science and would like to identify trends. Basically, your workflow already covers our task of analyzing the hashtags of tweets and creating a tag cloud based on that. That works for us.
Now we want to filter and display not only the hashtags in the tag cloud, but also the most important words in the tweets in general. These words should then be displayed in the tag cloud. I have attached a screenshot of the current tag cloud, which still needs to be cleaned up a lot.
Maybe you can help me with my problem and how the workflow could look like to clean up the data.
