Description
Once the list of comments was set up, we could read them all; with a further selection of the ones we could have used for a mood analysis, we excluded all the comments that did not contain opinions (ex. person tag, irrelevant emoji). We divide them in main categories (angry, contemptuous, critic, provocative, resigned, worried, ironic, supportive and skeptical), creating the second visualization; it shows whether the users agreed or not with the shared content in their comments. It was particularly interesting to find out that there were two main positions: the ones who were agreeing with the shared content and who expressed angry comments especially towards the government and the avocado mafia cartels. On the other hand several users were really skeptical about the news, stating that everything that was said was not true, that was a shame and that someone was trying to ruin Mexico’s image. In this way, the visualization resulted to be a sort of spectrum where we placed all the users’ comment mood stating if they agreed or not.
Protocol
Starting from the information listed in the previous protocol, we designed this visualization showing out analysis of comments’ mood. We first excluded the ones that could not have any mood assigned, such as the ones that were used to tag a person or an emoji, not expressing any particular feeling. We calculated the amount of every emerged mood (angry, dispregiative, critic, provocative, resigned, worried, ironic, supportive and skeptical) and we used this number to define the circles’ size, as wide as the related amount of comment mood. Placing these clusters on an “agreeing spectrum”, we could show whether that mood was near to agreement with the shared content.
Data
Data Source:
Google advanced RAWgraphs Statista Mexico Social Network Penetration
Timestamp: 19/10/2018
View Dataset
With the information collected for the previous visualization, we read all the comments, translated them in order to categorize the mood.