It was found that many people in Flint, Michigan were posting updates on social media (namely Twitter). What’s wrong with that? Anybody can post on social media. You may be familiar with the scandal in Flint where their water supply had more than the safe limit of lead in it. Some people in the city were actually voicing their concerns about their water, saying things like, “the water smells” or “the water tastes bad,” respectively. Some also posted with a location tag so everybody can see where they were when the water was smelly and it tasted bad. This is not the worst part, though. People were complaining about these things for more than a year before it was declared that the city was in a state of emergency which means that they kept using the water despite their concerns being broadcast all over the internet.
The time lag between concerns and public response in Flint was the main reason for the inception of the idea to mine social media data as well as search trends in order to understand what the public is experiencing in order to limit the time lag between the identification of a problem and the action taken as a response.
The idea turned into an environmental monitoring system which is essentially what its name implies, it tracks social media postings based on location in order to identify potential environmental hazards. This does not mean that all of our posts and searches will be monitored. It simply means that if someone posts something that could imply an emergency for the whole city or town, then the people who are behind the whole system will wait to see if more people are also posting about the same thing and then take the action that is deemed necessary.
The idea is credited to Pooja Chandrashekar, a biomedical engineer. Teamed up with his classmate Jeremiah Liu and a public health doctoral student, Yulin Hswen, he pushed the idea into a startup and won the McKinley Social Grant which won them 10,000 dollars to fund the Planetary Health Watch, as the team calls their startup.
The team has already collected approximately one million data points containing information related to water and pollution from Twitter exclusively. "When people talk about their concerns on Twitter, for example, many of the tweets will be geocoded, or users will self-identify the locations in their tweets. If you see a cluster of people tweeting about the water smelling really bad, and they are all concentrated in one area, you can become more confident that there is a problem there." Chandrashekar disclosed.
The biggest challenge for the team currently is developing the algorithms necessary to sort out those data points based on relevant data. They need to find out how to find people who are concerned and how to leave out those who are not posting any pertinent information. For example, if someone is posting about an organization to raise awareness for pollution and the system is looking for the keyword: pollution, the poster will be pooled in with people who are posting about pollution in their town that can actually harm them and others in their town or neighborhood or city. So the team need to figure out how to leave that person’s post out of the pool of posts.
The team does not expect their system to replace the EPA, however, as that would be a near impossible feat. Instead they hope for their systems to be incorporated into the EPA since the manpower would be significantly greater as well as the pay.
The thing that people posting need to remember if they are going to complain over the internet is to make sure that they add their location on the post so the system can find that particular post. Leave the rest to the computers.