HashtagHealth-Geoportal

About the Geoportal

This online mapping application may be used to look up geotagged Twitter data on indicators of happiness, diet, and physical activity at the state, county, zip code or census tract levels. *Twitter data collection took place between April 2015 and March 2016. The data dictionary includes descriptions of constructed variables. Data summaries were derived from 80 million tweets from the contiguous United States. Please refer to the following manuscript for more details on the data collection. The mapping application was built with Carto and Google Maps

The specific functions of the Geoportal map are as follows:
Function 1. View data at the clicked location.
Function 2. View data at the searched location.
Function 3. View aggregated data at a drawn region by choosing the "Aggregate" button after drawing the region with the tools on the top of the map.
Function 4. Download raw data at the state, county, zip code, and census tract levels in csv or txt format with links in the window.

By Minh Pham and Matt McCullough
Principal Investigator (K01ES025433): Quynh C Nguyen

Geoportal

Open new tab for a full view of the mapping application

Google Street View

Dataset

View Google Street View data

About Google Street View data

We utilized Google’s Street View Image API to collect image data from Chicago, Illinois; Charleston, West Virginia; and Salt Lake City, Utah between December 2016-February 2017. Details on data collection and image processing are included in a separate data dictionary. This dataset provides zip code and census tract summarized data for the following neighborhood characteristics: 1) street greenness/landscaping (street trees and street landscaping comprised at least 30% of the image; yes/no), 2) building type (single-family detached house vs. other), and 3) presence of crosswalks (yes/no).

For each zip code, the following was calculated: percentage of street images with A) presence of a crosswalk, B) building type other than a single-family detached house, and C) street landscaping comprising at least 30% of the image. Neighborhood characteristics were categorized into tertiles and spatially mapped. Darker colors represent higher tertile values. Data source: Google Street View Images.

Figures

Figure 1. Zip code distribution of built environment characteristics in Salt Lake City, Utah

Figure 2. Zip code distribution of built environment characteristics in Chicago, Illinois

Figure 3. Zip code distribution of built environment characteristics in Charleston, West Virginia

People

Faculty

Dr. Quynh Nguyen (Principal Investigator), Assistant Professor, Department of Epidemiology and Biostatistics, School of Public Health, University of Maryland, USA

Research Assistants

Matt McCullough, Ph.D. Student, Department of Geography, University of Utah, USA

Minh Pham, B.S. Student, School of Computing, University of Utah, USA