Advertisers understood it a long time ago: social media is an absolute gold mine of data. Today, as the corona crisis continues to rage across the globe, social media data can, more than ever, be used to drive action. Academic researchers and governments are therefore turning to social media to collect huge amounts of information and gain essential insights on the coronavirus outbreak, its consequences, and how to deal with them. Here are two examples of ongoing data-mining projects in the Netherlands and beyond, that combine artificial intelligence and social media to guide decision-makers.
The PuReGoMe project, an initiative of Utrecht University and the RIVM to study public opinion in corona time
Utrecht University is partnering up with the RIVM (Rijksinstituut voor Volksgezondheid en Milieu, the Dutch National Institute for Public Health and the Environment) to conduct a nationwide study via social media, with as primary goal to investigate public opinion about the corona-related measures.
Understanding how the public feels about the government’s actions in this time of crisis is fundamental: it can help policy-makers take more adapted, enforceable decisions and even predict whether or not people will adhere to measures. Are people confident, trusting that the government is doing a good job in protecting them? Or are they resigned, pessimistic and sceptical? Is there a worrying climate of fear, or of resistance, slowly creeping in?
These are things the Outbreak Management Team and the prime minister want to know, so that they can address the issues, proactively. These are things our social media posts can tell them. Indeed, each time new measures intended to “flatten the curve” are announced, the reaction of the Dutch public is rapid, clear and mostly expressed on… social media, and more particularly, Twitter.
The PuReGoMe project, or “Dutch Public Reaction on Governmental COVID-19 Measures and Announcements”, is funded by the eScience Center and will last a total of six months, starting May 2020. You can read more about it here.
The gist of the project is quite simple: the researchers select Twitter data on a particular trending topic (in this case, corona measures, corona statistics and corona news) using relevant hashtags and keywords, and then use artificial intelligence and natural language analysis to find out whether people have a positive or negative response. They can run case-by-case analyses to study the public’s response to a specific measure, or they can monitor changes in opinion and new reactions following a specific event (when RIVM publishes new figures, for instance). Of course, Twitter data alone is not necessarily reliable, which is why the results are also cross-checked and corroborated with data from other social media.
The University of Maryland corona-mapping project
The data collected on social media can be used for much more than only gauging public opinion. For example, the spread of the virus, the infected “hot spots” and the most economically affected areas can all be mapped using social media. It goes without saying that this information must always be taken with a pinch of salt. Nonetheless, it gives a good enough overview of the situation in real-time, which is a non-negligible advantage in our fight against the coronavirus.
You may have answered it yourself recently, the University of Maryland is conducting a worldwide survey on Facebook to try to forecast possible new outbreak areas, and to identify the regions that need extra support, be it sanitary, financial, psychological or social. They ask questions such as “how worried are you about being able to eat for the next month?” or “how often do you wash your hands?”, “how many people outside your household have you been in close contact with (less than 2 meters distance for more than one minute) during the past seven days”, “have you been tested for COVID-19?”, “do you know anyone who has?”, etc.
Only future will tell how much all of this gathered data really helped, but the super-fast communication between everyone that is made possible by social media is without a doubt a huge asset when it comes to improving our preparedness in this global pandemic.