Harnessing social media to track potential disease outbreaks
Channel NewsAsia looks at how social media can provide reliable data in time of crisis, and how information found on various platforms can be used to track potential outbreaks locally.
- Posted 30 Dec 2015 23:22
SINGAPORE: Preliminary views from the taskforce looking to enhance the response to infectious disease outbreaks in Singapore included better use of data analytics and IT systems.
Data analytics involves the examining of raw data to discover patterns and information.
With the proliferation of online networking sites, a team of Singapore researchers said social media can provide a rich source of reliable data in times of crisis, and it is looking into harnessing such information to track potential outbreaks locally.
Singapore researchers were prompted to embark on a study of the bird flu outbreak in China back in 2013 when asked if social media platforms would be able to help in battling the spread of an infectious disease like bird flu. Results of the study were published early this year.
The researchers collated information from Chinese social networking site Weibo and compared it with the information streaming in from traditional sources such as the Chinese health authorities, the World Health Organization and news websites.
The team found that social media could indeed provide reliable and timely information in times of crisis. The team managed to track the spread of bird flu in China and pick up first-hand reports about the disease by trawling cyberspace.
It was revealed that Weibo was significantly faster in reporting new cases compared to conventional public health channels. On average, new cases were reported on Weibo about an hour before the Chinese health authority's website.
"We found that the contributors to the early reporters were actually the local news agencies. They have dedicated personnel to post on Weibo even before they release (information) to conventional newspapers, print media,” said Dr Yang Yinping, Capability Group Manager, Social Intelligence, Institute of High Performance Computing.
Information from the man in the street also provided useful epidemiological information about the disease, such as patient symptoms or lifestyle history.
"For example, there was actually a patient who is a vegetarian. That means this person shouldn't eat chicken, but he still got infected. So this is important lifestyle information for epidemiologists and clinical professionals to take into account,” said Dr Yang.
Dr Yang added that the system can also help health authorities assess public sentiment to formulate policies.
But with a barrage of information online, researchers also found it a challenge sifting through what was real and what was not, so filtering techniques were used to verify information.
The team is taking its research one step further. Over the past one and a half years, it has been working on an almost-real time information gathering and response facilitation system that can help authorities in Singapore track any potential disease outbreak.