1. Mapping the (Digital) Moral Domain: Using
Mined Twitter data in investigating the
Sociology of Morality with Undergraduates
Dr. Brian Lowe, SUNY College at Oneonta
Mr. James Greenberg, SUNY College at
Oneonta
Digital Sociology Mini-Conference (2016)
@ The Eastern Sociological Society
2. Why Investigate the Sociology of Morality with
Social Media?
• Objectives of introducing undergraduates to social media data
gathering and analysis:
• Introduction to data-gathering
• Hypothesis testing and theoretical application
• Infusion of academic materials with unfolding events
• Excellent grounds for exploring constructivism
3. Exploring Constructivism with Social Media
• Investigation of terminology used in primary sources
• Confronting the ambiguity in terminology used in social media
• Assessing apparent patterns in data
4. Case Study: Cecil the Lion
• 13 July 2015: Cecil is killed near Zimbabwe’s Hwange National Park
• 28 July 2015: UK Telegraph identifies American Dentist Walter Palmer as
hunter who killed Cecil
• Reaction in mass media and social media is swift
• http://www.telegraph.co.uk/news/worldnews/africaandindianocean/zimb
abwe/11767119/Cecil-the-lions-killer-revealed-as-American-dentist.html
5. Case Study: Cecil the Lion
• 4 August 2015: Image of Cecil the lion projected onto the Empire State building
• Legislative action
• 16 September 2015:
• HR. 2494 Global Anti-Poaching Act
• S. 1918 C.E.C.I.L. Act
• 2 November 2015: H.R. 2494 passed with bipartisan support
• http://www.takepart.com/video/2015/12/14/how-cecil-changed-wildlife-policy
6. Case Study: SeaWorld
• 24 February 2010: Tilikum (Orca) kills trainer Dawn Brancheau during public show at SeaWorld
• 23 August 2010: OSHA fined SeaWorld $75,000 for violations related to Dawn Brancheau’s death
• 2013: Blackfish released (directed by Gabriela Cowperthwaite)
• Made for $76,000 US dollars; initially screened in five theatres
• Aired in October 2013 by CNN to an estimated audience of 21 million
• http://www.nytimes.com/2010/08/24/us/24brfs-SEAWORLDFINE_BRF.html
• http://www.cnn.com/2014/02/07/us/blackfish-wrap/index.html
7. Case Study: SeaWorld
• 2013-2014: Withdrawal of celebrity entertainers including
• The Beach Boys
• Trisha Yearwood
• From “Bands, Brews and BBQ” event
• August 2014: SeaWorld stock had lost more than 30 percent of previous
value
• http://www.nytimes.com/2014/08/14/business/media/media-scrutiny-
drags-on-earnings-at-seaworld.html
Editor's Notes
This is a screen shot of the CecilTheLion data that Diana Moseman’s process captures. This data comes in as JSON formatted. We zip up all the data files and transfer them to VIDIA for conversion to a CSV. This data can be processed on VIDIA using tools like Rapid Miner or R. Or we can transfer the CSV back to our local computer and process it using Modeler. Or we can upload this data into Watson for analysis.
This is that same data set as show in the previous slide processed with Rapid Miner on VIDIA. The graph shows word counts for any word that appears more than 6000 times in the data set (pruning). The data were tokenized, stop words were removed as well has some other noise like “http” of URLs.
This is a screen capture of an IBM Watson generated dataset on Cecil the Lion. It shows how many occurrences of the topic appeared each day between the Dates selected (see above in the Dates). There is some discrepancy here between what is being show and what is selected that has to be investigated. This often happens and is a good process for students to wrestle with.
This is a more sophisticated analysis of two concepts plotted together to see if any time relationships exist. Again, note that anomaly in the data near the right most part of the X axis. At first glance one would suspect a data error of some type. This would have to be investigated and explained. Note how some of the high volume days seem related.