Nowcasting Obesity in the U.S. Using Google Search Volume Data

“Googling” is now ubiquitous in our society. We typically start searching on Google before we make purchase decisions or when we have interest in certain topics. Aggregating these search data can provide us with real-time and possibly accurate information on people's behavior. In fact, Google keeps tracks of all the search queries and has accumulated a tremendous amount of information about people's interest at the society level. It currently provides search volume data of keywords for different regions and time intervals on its free and public service of Google Trends. An interesting and hot research area is how to exploit the Google Search volume data in innovative ways to benefit our society. This paper aims to reveal the connection between obesity prevalence and people’s online search behavior in the United States by combining data from Google Trends and data from Behavioral Risk Factor Surveillance System (BRFSS) which is updated by the Centers for Disease Control and Prevention (CDC) annually. We first handselected keywords that are associated to people's life style and used panel data model to study association between search pattern and obesity level. We found significant correlation power of those keywords with Body Mass Index (BMI) level and results suggest great promise of the idea of obesity monitoring through real-time Google Trends data. We believe this is an important finding and is particularly attractive for government health institutions and private businesses such as insurance companies etc.


Issue Date:
2014-06
Publication Type:
Conference Paper/ Presentation
PURL Identifier:
http://purl.umn.edu/166113
Total Pages:
33




 Record created 2017-04-01, last modified 2017-08-27

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)