The workshop this afternoon on mining and analyzing social media content from Twitter and Facebook was fantastic!
Mining Twitter content, we calculated consumer sentiment, created pie charts, word clouds by sentiment, and LSI topic modelling by sentiment. (Ouch, apologies for the adult words that snuck into the negative sentiment tweets!)
Mining Facebook, we calculated conversation, applause, and amplification rates by month.
The website for the workshop, including all Jupyter Notebooks we used, is available here: https://researchguides.baylor.edu/c.php?g=813695&p=5806024
Carol started us off giving an excellent background on the topic and then I guided everyone through the exercises. With a packed house filled with undergraduate students (thanks Dr. Randy Hacker for offering extra credit for attendance!), we had such a great time engaging them in hands-on social media data analytics.
We were able to highlight some of my favorite Python libraries, including Gensim and VADER Sentiment.
This workshop generated a potential repeat performance, request for an NVivo workshop for social work faculty, data transformations support for management faculty, and a horde of undergraduate students asking for assistance installing and learning to code in Python. Fantastical day!