Description
In the visualisation you can see the reactions we extrapolated from the comment section of diverse video about self driving cars safety. We can see how in a relatively non-controversial context, the strongest sentiments present seem to be the analytical one. Tentative, irony and anger also seem to be strong and fluctuant over time. What emerges is an intense but yet mature debate around a subject which is perceived as highly controversial for the modes in which it’s being tested on people while the technology isn’t fully developed at the moment. References to the Tesla’s car crash are also present, outlining once again the key role of these driver-events.
Protocol
Using the dataset of 82 videos that we obtained on Youtube channel in wich we analyzed comments. Selecting all of the comments that have more then 1 like. In the end of scraping dataset we were left with 6773 comments. We used two methods of analyzing comments to get more reasonable resaults. We used the IBM ToneAnalyzer, a tool which assigns an emotional score to words in order to obtain the general sentiment of a public by processing big datasets. From a previous experience analyzing sentiment and polarization on 500 newspaper articles comments we knew we needed to use a much higher number of comments in order to obtain a reliable and legitimate score. We also elaborated a protocol to assign a “manual” value and also by adding “irony” as the new possible feel which we perceived was highly relevant to due to communication modes of today’s web. The feeling analyzed were joy, sadness, anger, fear, confident, surprised and analytical, plus the irony for our manual score. We then elaborated the data into our visualization in Illustrator.
Data
Timestamp: 01/12/2016 - 05/12/2016
Data source:Youtube
Download data (4MB)
The dataset is an .xls file with the list of comments, links, text, and the score assigned to the different emotions by the IBM tool and us divided in seven/eight categories: joy, sadness, anger, fear, confident, surprised and analytical, plus the irony for our manual score..