How Big Data and AI can help in fighting COVID19
As the panic continues to spread worldwide and the COVID-19 continues to infect more and more people, the first measures to contain the virus are beginning to be seen: frozen businesses, blocked transport, closed kindergartens, schools and high schools or canceled events are some of the examples. Every country is responding to virus propagation by taking the measures it considers appropriate and trying to break the virus spread inside and outside their frontiers. Over the next few weeks, the progress of the virus’s expansion must be closely watched by governments and they have to check if the measures that are being taken are enough but, for now, the virus spread is inevitable. More and more infected persons are detected and new countries are being added to the list of the already affected. If you want more detailed insights into the COVID-19 status, you can check up the last information on the European Centre for Disease Prevention and Control website. Due to the global situation we’re in, I would like to analyze how Big Data and Artificial Intelligence can help, not only in this specific case of the COVID-19 but with a broader perspective in all disease control situations.
Both the movement of people that occurs within a country and between borders is increasing every day. Working needs, cheaper transportation, and tourism itself promote the continuous flow of people. This, added to the high density of some countries, such as China, generates an environment where it is very easy for a virus that is transmitted through the air to spread. Due to this continuous flow of people and the moment of globalization in which we are, it is not possible to delimit this environment within our borders, but an approach must be made with a global perspective: we have a global pandemic.
What is the use of Big Data and AI on disease control?
The continuous movement of people that I introduced before, is completely measurable and with those measures, we can build a set of data that follows the 5 dimensions of Big Data (if you don’t know what I’m talking about, I suggest you read my previous article) and extract value from it. But, where do those data come from? I defined a list of some data sources that can provide enough information to perform a future analysis or create valuable services for governments or citizens:
- Geolocation of mobile devices. In a first impression, you can think that this is a little bit dangerous and violates people’s privacy rights, but, if we think carefully about it, private companies like Google or Facebook are already doing it with our permission for commercial purposes. If you don’t trust me, check by yourself. If you have a Google account, check this page or, if you are an Apple fan, take your iPhone and have a look at:
Settings > Privacy > Location Services > System Services > Significant Locations.
- The information reported by citizens through official channels. Countries like South Corea developed applications to enable a fast communication channel with citizens. If they have any symptoms, they can report them through this application, request a COVID19 test and follow their situation.
- Social media and news. This point is a little bit tricky because not all the published information has the same veracity grade, but have a look at Google Trends. You’ll see that words like “coronavirus” or “pandemic” are on the top of the list. Also, “COVID19” is a trending topic on Twitter.
- And, finally, we cannot forget the COVID19 detection tests. Another possible source of data is the record of all the results of the COVID19 tests that have been carried out all over the world.
Once we have identified some of the possible data sources, let’s think about the uses that can be made of that data. First of all, I would like to clarify a point about privacy. Even though part of the data collected in the previous section may threaten privacy (especially the results of the coronavirus tests), we must emphasize that the treatment of these data must be previously anonymized, that is, it is not about identifying specific people, but infected and uninfected.
Trending AI Articles:
1. Bursting the Jargon bubbles — Deep Learning
2. How Can We Improve the Quality of Our Data?
3. Machine Learning using Logistic Regression in Python with Code
Anyway, although some of the use cases that I will propose could be carried out with public data, these examples have been thought of as the uses that governments could make of that data (some have already been carried out in certain countries).
- First and foremost, visualize and explore the past is the best way to understand it. It is very difficult to understand and compare numerical values, and displays, whether based on colors, magnitudes, or maps, for example, help to understand the meaning of numbers. In that sense, I recommend you take a look at the dashboard published by the NYTimes.
- Knowing which people are infected, when infections have been detected, where they have passed, where they have traveled, and by cross-referencing that data with where other people have been or traveled, the spread of the virus can be predicted with a certain degree of precision. Besides, interesting insights can be enabled, for example, to measure virus spread rates.
- Calculate the level of risk of contagion of a person based on their activity and take preventive measures. Many business models are beginning to use preventive maintenance to predict when their systems will need maintenance. If we extrapolate this example to the case we are analyzing, we can do “predictive maintenance” of people and anticipate their possible infection.
- One of the biggest problems in these situations is to manage public opinion. People, in the face of misinformation and uncertainty, point to overwrought and overreaction. Furthermore, the population is in a continuous flow of information from social networks and news that can be false or ambiguous cause more confusion. In this regard, governments can develop conversational bots that interact in natural language with persons and provide them the latest public information or answer to specific questions.
- Another use case of artificial intelligence is facial recognition. For example, in our modern mobile devices such as iPhone or Samsung, the unlock feature is developed over an artificial intelligence engine that can distingüish our face from any others with a very high success rate. Think in a drone that can apply this artificial intelligence technique and detect if an infected person has breached its quarantine, it sounds like science fiction, right? Maybe, but China is already doing it.
These data sources and use cases are just examples of the many applications that can be carried out. Even though Big Data and Artificial Intelligence are increasingly present in our day to day, my intention with this post only is to illustrate how they can be of great help to the social environment and, more specifically, in disease control and prevention. If one thing is clear, it is that we are not prepared for this type of situations and, from my point of view, we must work hard on this aspect since it is not difficult for a new virus or a mutation of an existing one to affect us within some years. We are living in a situation of uncertainty and confusion and what is clear, although it is still pending to determine, is that there will be economic consequences at a global level: the world has stopped and the stock markets are falling day after day.
Don’t forget to give us your 👏 !
