Revolution #5: Say Hello to the geoLab’s Newest Addition: geoParsing

Matt Crittenden is the Team Lead for geoParsing | Email:

In November 2018, the geoLab began planning the introduction of a new fourth team, geoParsing, to solve two serious problems.

Why Create a New Team?

In the past, academic research labs (like AidData at William and Mary) have serviced organizations around the world by locating and assigning spatial data to international development projects and financial activities. However, organizations and individuals unable to afford these services were often left out. The first purpose of the new geoParsing team is to resolve this issue by providing non-time-sensitive deliverables to organizations previously unable to access these services. The second purpose is to create a methodology for tracking financial activities which can be easily tailored to unique country contexts.

What does geoParsing do?

Our current project is ambitious and challenging. We are working with a partner institution to track a country X’s financial flows and activities in a Sub-Saharan African country Y. Existing projects have done little tracking specific to X, so this project presents an exciting opportunity to define our new methodology. Unlike previous projects, we do not have documentation on where funds are going or why. Starting from square one means literally going online and figuring out what’s going on with X’s activity in Y, then identifying reliable sources of information, parsing through and geocoding all relevant financial information, and, finally, double-checking that our process was sound.

It’s going to be a learn-as-we-go process. To boost efficiency, we will develop and employ new tools. We just launched version 1.0 of “GeoHighlight,” a program which scrapes documents for relevant location data. This will help RAs identify the project locations to geocode. Soon, we will explore creating a web-scaping tool to locate relevant sources and project documentation on the web.

Ultimately, the goal of this project is to find and give coordinate data to all of X’s financial activities in Y. The future is busy, bright, and promising for geoParsing!