Researcher Says Twitter Holds the Potential to Alter Public Health Policy
BY Lisa Goldman | Wednesday, September 19 2012
Twitter holds the potential to yield valuable information that could alter the way data about public health issues is collected and analyzed, according to a recently published academic study.
Mark Dredze, an assistant research professor of computer science at Johns Hopkins University, wrote a research paper called How Social Media Will Change Public Health, in which he discusses how he has used an algorithm to mine tweets for information about self medication, exercise, dental pain and smoking to augment and create new public health capabilities.
To explore these tweets, we developed the Ailment Topic Aspect Model (ATAM), a probabilistic graphical model for uncovering ailments.
ATAM assumes that each message discusses a single ailment, manifested through the message’s words, and associates three types of words (general disease words, symptoms, and treatments) with ailments. For example, the message “fever + headache = flu, home sick with Tylenol” discusses influenza, where “fever” and “headache” are symptoms, “Tylenol” a treatment, and “flu” a general word associated with the ailment.
Human annotators labeled 15 ailments discovered by ATAM, including headaches, influenza, insomnia, obesity, dental problems, and seasonal allergies. Examining the words, symptoms, and treatments most associated with each ailment, and the groups of messages that discuss each ailment, can support a variety of public-health initiatives.
Dredze goes on to provide several concrete examples of how ATAM can be applied to cull valuable information from millions of aggregated tweets — e.g., to track allergy season or to assess risk factors of conditions like obesity and a sedentary lifestyle. He also shows how analyzing data obtained from tweets can be a more efficient means of assessing populations and developing policies.
These successes [of Twitter's demonstrated ability to deliver valuable information in real time] have drawn interest from the public-health community, whose goal is to study the health of a population and develop policies that improve health outcomes. Traditionally, this requires expensive, time-consuming monitoring mechanisms, primarily surveys and data collection from clinical encounters. Even high-priority projects, such as the US Centers for Disease Control and Prevention’s (CDC’s) FluView program that tracks the weekly US influenza rate, are still slow because they require clinical data aggregation. Twitter and other social media could reduce cost and provide real-time statistics about public health.
Dredze and one of his students explain how ATAM works in this very cogent video presentation of only one minute and 40 seconds:
Dredze acknowledges that clinical research on the effectiveness of Twitter as a public health tool will have to account for bias — e.g., the fact that it is not representative of the entire population, or that people can report inaccurate information. "Controlling for bias is a hallmark of clinical research," he points out. But the biases of social media are not yet fully understood.