Automated Web-based Archive System for Local School Wellness Policies
- Principal Investigator
- Chriqui, Jamie F.
- Start Date
- End Date
U.S. school environments are important battlegrounds in the fight against childhood obesity. Literature suggests that effective school wellness policies can improve children’s dietary intake and levels of physical activity. Therefore, federal regulations require local schools to develop and implement wellness policies. Moreover, the Healthy, Hunger-Free Kids Act of 2010 requires that schools periodically update on the progress made in the implementation of their school wellness policy and that federal and state agencies should monitor school wellness policies for compliance. Therefore, there is a need for a surveillance and searchable system for school wellness policies.
However, due to the large number of school districts and various policy formats used, monitoring changes in these school wellness policies presents a daunting task for government agencies or researchers. The traditional annual survey of sample schools is inadequate and not time sensitive to meet the challenge. The USDA and National Association of School Boards of Education (NASBE) openly call for innovative methods to monitor school wellness policies at the national level. To respond to the significant needs, we propose to develop an automated web-based archive system to harvest school wellness policies online.
Our system will automatically search large numbers of web pages to monitor policy updates, harvest school wellness policies from web sources which are not accessible to direct search engines, catalogue these policies in a searchable database, and provide an interface to access and monitor wellness policies.
Traditional harvesting and evaluation of health policies are human-oriented, which could be labor-intensive and costly. Our long-term vision is to substitute human efforts in this process with artificial intelligence, offering higher efficiency. This pilot project is the first step in an exciting journey to use machine learning in school wellness policy harvesting. If successful, this system will not only be a useful tool for government agencies and other stakeholders to monitor school wellness policies, but also prove the feasibility that artificial intelligence can complement human efforts in health policy harvesting. Researchers can utilize the experience and knowledge obtained from this project to develop harvesting systems for other public health policies.