Jonathan Kropko wants to help build something special at the University of Virginia’s new School of Data Science, and just as important, he wants his work to have a positive impact on the Charlottesville community at large.
So when the opportunity to help launch Code for Charlottesville presented itself, the UVA assistant professor jumped.
The nonprofit, volunteer organization – which includes volunteers from UVA, the city and surrounding region – is part of Code for America, founded in 2009 to address the widening gap between the public and private sectors in their effective use of technology and design.
“Code for Charlottesville is a way for people to do work that has a direct impact on their community,” Kropko said. “It’s a way for talented people to use those talents for community service.
“It is my sincere hope that Code for Charlottesville helps break down longstanding divisions between the University and the town. We are an organization that brings people together to work toward common causes.”
Kropko, who has a background in political science, math and applied statistics, had attended Code for Durham and Code for Raleigh events in North Carolina and wondered about the possibility of starting a Charlottesville brigade.
When he returned to Charlottesville, he found that UVA fourth-year student Elyse Lee was already doing so. So Kropko came aboard as a co-captain of the organization, which launched in September.
The group’s first project was with Charlottesville’s Legal Aid Justice Center.
A new project with the Charlottesville Office of Human Rights that will help people with limited incomes find places to live launches Tuesday evening. Kropko said anybody interested in joining the team is welcome at the launch event.
UVA Today caught up with Kropko to learn more.
Q. What is the goal of Code for Charlottesville?
A. Code for Charlottesville brings volunteers with skills in programming, project management, data science and design together to work on community service projects with local nonprofit and governmental organizations. We also want our events to be good places for people to meet each other and to get exposed to new ways of doing community service with a technical component. As Charlottesville grows and its tech industry gets healthier and bigger, we want to be a venue where people with skills can apply those skills for the benefit of the community.
Q. Some people might see the word “code” in the organization’s name and assume you have to be some sort of computer whiz to be a part of the team. Is that true?
A. We need lots of help that isn’t just coding! We need designers, statisticians, project managers, writers, policy students, community organizers, translators.
Our project with the Legal Aid Justice Center, for example, involves building a website to educate people about their rights in Virginia with regard to wage theft. The work involves code to create the website and the interface to guide people through the process of filling out a government form. But it also involves user-experience design to make the website as user-friendly as possible; it involves writing to communicate the information effectively; it involves Spanish-language translation, so that we can speak to more people who are affected by wage theft; and it requires a thorough understanding of the problem of wage theft and the current laws that people can use to fight for their rights.
Q. For the non-techies out there, can you explain what a “hack” is?
A. Lots of things are called “hack” these days. We use the word to describe our events in which people break off into teams to work on some part of the larger project. Hacking is just focused time to do work. (It’s also a pretty cool word! We get bigger turnouts at “hack nights” than at “work nights.”)
Q. You’ve already had a couple of big projects. The first one was helping the Legal Aid Justice Center with wage theft in Virginia. How did you land on that for your first project?
A. A team at the Legal Aid Justice Center, headed up by Nicholas Marritz, contacted the School of Data Science to request help cleaning and analyzing data that they got by issuing a Freedom of Information Act request of the Virginia Department of Labor and Industry. DOLI’s data contained information on every official complaint of wage theft they received over the last five years. I work at the School of Data Science, so I saw the opportunity for Code for Charlottesville to get started with this great project.
The DOLI dataset was in pretty rough shape, with formatting errors, inconsistent codes and misspellings (several cases were from “Rochmond”). In our first three hack nights, Code for Charlottesville cleaned the data and created maps and tables to summarize the data. This work helped the Legal Aid Justice Center write a report on wage theft that is currently getting a lot of attention in state government, and hopefully will lead to some positive changes!
Q. What were the challenges you faced in trying to come up with a solution?
A. By far the biggest challenge is finding a way for a large group of people with many backgrounds to work together in an organized way toward an overall goal. Crowd-sourced work is difficult because people might duplicate each other’s work, because it’s hard to communicate what needs to happen and when, and because it’s tough to both onboard people onto a new project and at the same time allow returning volunteers to build off of previous work.
Q. Code for Charlottesville’s other project has been with the Charlottesville Fire Department. Can you give us a brief snapshot of what you’re trying to do?
A. The Charlottesville Fire Department wants severity-risk scores: numbers that indicate how severe that fire is likely to be, given things we know about where the fire is, such as the number of buildings on that parcel, the age of the buildings, the building material, and so on. The idea is to make these scores available to the 911 dispatcher office as well. That way, if there’s a call about a fire, the dispatcher will see the severity-risk score on the screen and use that information to choose how many fire trucks to dispatch. The first responders on their way to the fire can use that score to determine what equipment to prepare.
Code for Charlottesville held hack nights last November and December on this project, and is continuing the work now with a smaller, dedicated team. The work involves merging many public datasets from the Charlottesville Open Data portal together, matching it with historic fire data, generating the scores, plotting them on maps, using machine learning to improve the predictive accuracy of the scores, and building an appealing user interface for the Fire Department and the public to see these scores.