Imagine knowing exactly when queues are going to form at medical centres and hospitals. And, even better, to know what treatment these people need before they are even examined.
Health managers could then plan on when to have more doctors rostered, what sort of drugs should be on the shelves and what type of equipment is needed.
Predicting the future has been an industry for thousands of years. The benefits are obvious: when to plant crops; when to move and when to stay; where the risks will be and what they look like.
Until now, the art has been hit and miss. The rise of social media, whether through Facebook, Twitter or other platforms, has created an almost instantaneously updated pool of information about people, their thoughts, actions and feelings.
The trick is being able to analyse millions of messages, tweets and posts, to quickly assess the mood and health of the nation.
A study in 2011 by researchers at the University of Bristol’s Intelligent Systems Laboratory, and published in the journal ACM Transactions on Intelligent Systems and Technology, used geo-tagged posts on Twitter as an early warning system.
What 50 million tweets can do
The researchers looked at reports of flu-like symptoms, through the lens of the content of tweets. If people were talking about the flu or said they were showing symptoms that could be related to it, this got picked up, aggregated and analysed.
The idea was to see if an emerging epidemic could be uncovered faster than even local GPs and health centres because people were first talking about their symptoms online before seeking medical attention.
A database of more than 50 million geo-located tweets was created. This was then compared to official data from the UK’s National Health Service on flu incidence by region.
Human-only analysis was too frail to cope with such large amounts of data. Computer processing power and a series of algorithms judged which keywords were associated with a rise in the incidence of flu.
From this a model was created to predict how many people were about to, or were in the process of, coming down with the flu and how severe the illness was in a specific geographic area.
This geo-tagging of posts — labelling where the poster is — has led to a new wave of experimentation and research, according to Professor Nello Cristianini of the University of Bristol.
“Twitter, in particular, encouraged their 200 million users worldwide to make their posts, commonly known as tweets, publicly available as well as tagged with the user’s location,” he says.
“Our research has demonstrated a method, by using the content of Twitter, to track an event, when it occurs and the scale of it. We were able to turn geo-tagged user posts on the microblogging service of Twitter to topic-specific geolocated signals by selecting textual features that showed the content and understanding of the text.”
The current body of research using tweets to track illness rates led to the creation of the Flu Detector, shown here:
The detector, using Twitter feeds, is almost a real-time representation of the incidence of flu in several UK regions. While not everyone uses Twitter, there are enough users to make the results relevant.
In Australia, there’s been research on judging the mood of the population as part of work to define and treat mental illness.
Public health programs are mostly created based on research and statistics at least five years old. However, social media can give an instant picture of what people are feeling now. By sizing the problem, estimating the extent of mental illness, policy makers can plan to ensure the right resources are allocated to meet the problem.
Here’s what the output from the tool looks like:
“It uses language-processing techniques to look at the English words people use in these posts and then maps these words to a hierarchy or wheel of emotions,” the CSIRO says.
“We Feel allows you to explore emotions visually across a minute-by-minute time scale which extends back days or several weeks. Users can also explore emotions across locations around the globe and select other search criteria such as gender to further refine the results.”
The tool also helps to judge how emotions depend on social, economic and environmental factors such as the weather, time of day, day of the week, news of a major disaster or a downturn in the economy.
Another way to use social media is to track food borne illnesses, many of which are just not picked up by health authorities because few people seek medical care or report to local authorities unless they are so sick they need an ambulance.
In the US, biostatistician Elaine Nsoesie wanted a surveillance tool to help local public health departments with food poisoning outbreaks.
“Online reviews of foodservice businesses offer a unique resource for disease surveillance. Similar to notification or complaint systems, reports of foodborne illness on review sites could serve as early indicators of foodborne disease outbreaks and spur investigation by local health authorities,” says Nsoesie. “Information gleaned from such novel data streams could aid traditional surveillance systems in near real-time monitoring of foodborne related illnesses.”
The next step is to develop a dashboard for processing reports of foodborne illnesses for the purpose of further investigation and restaurant inspection.
“This data-mining framework would hopefully provide near real-time information on foodborne illnesses, implicated foods and locations,” she says.
Researchers have also created an app, which scans social media and news reports, to assess the risk of an EBola outbreak.
The idea is to create a running score assessing the level of danger from Ebola when travelling.
“Our feeling regarding Ebola is that there is a high level of anxiety due to a lack of experience and education,” says Jun (Luke) Huan, professor of electrical engineering and computer science at the University of Kansas.
“Our app should help in both cases. It sends updated information of the virus to end users. It also computes a personalised risk score for the user so that the person is informed and updated.”
The app, called iChequIt, will alert if the user goes to an area that a newly reported Ebola patient visited.