Big Data

Meet Data for Good, the Hacker News for showing off your world-changing data science

From left, Data For Good founders Tobias Pfaff,  Vicens Fayos, and  David Geffroy.

Above: From left, Data For Good founders Tobias Pfaff, Vicens Fayos, and David Geffroy.

Image Credit: Robin Himmels
Got email marketing? We've got best practices from LivingSocial and estate sale guru Everything But The House in our next Insight webinar.

Data scientists already have DataTau, a specialized version of Hacker News, that popular online water cooler for programmers.

But some socially minded data scientists thought DataTau wasn’t great enough, and now they have their own site. Data for Good is a public-good-only fork of DataTau.

Data for Good, which itself is a nonprofit organization, represents a new community to display the most noble-minded efforts of data scientists as the field stays hot and more companies pursue data scientists.

Physical and virtual places to learn data science are multiplying, as are programs geared toward social good. So Data for Good’s appearance is timely.

Tobias Pfaff, Vicens Fayos, and David Geffroy built Data for Good 10 days ago, during a startup weekend in Hamburg, Germany. Pfaff hails from the German city of Münster. Fayos comes from Spain. Geffroy is French. Fayos and Geffoy work in Hamburg.

Pfaff, who lives in Munster and took a Ph.D. in economics from the University of Münster last year, wondered what he could do with predictive analytics after being inspired by a big data book that talks about how the New York City Mayor’s Office of Data Analytics prevents fire risk with predictive analytics.

“So I thought, OK, I do predictive analytics, maybe I can replicate their cases in Germany,” said Pfaff, in an interview with VentureBeat. “I approached the state office for food control. I wanted to use predictive analytics to make their processes more efficient, because we have a problem here in Germany with too few people doing the food controls.”

The state office, which dispatches inspectors to check out restaurants, farms, and animal feed producers, was very interested at the beginning, but at the end of the day, it wanted to wait.

Nevertheless, Pfaff continued to collect data analytics use cases around the web and compiled a list of projects for himself. “I always thought, why don’t we have a centralized directory for such use cases?” Pfaff said.

At the startup weekend, Pfaff pitched his idea and got Fayos and Geffroy together to bring it to life. They worked late into the night and launched a rough version of Data for Good after the startup weekend.

Data for Good aims to be a listing of plain records for all the socially beneficial data science projects. There are three types of people Pfaff wants to connect on this platform.

“People who provide the data, or the open data, they wanted these cases because they need something cool to be done with their data,” Pfaff said. “Civic hackers want these use cases because they want to be inspired by what they can do with open data. There are people at the administration [level] who are dealing with problems every day. They also need some solution to their problems.”

Pfaff is not trying to reinvent DataTau. The difference is that Data for Good is a list of socially beneficial data science projects, not something broad, so that people who go on the website can find work to replicate.

“Imagine someone from the local administration who is not really a hacker,” he said. “Will they go through all these post entries on DataTau to separate the project out from all the other data related stuff? I don’t think so. I think those people need a plain list of projects where they can see what’s happening.”

Pfaff has a long-term vision for Data for Good. The first step is a bulletin board. The second step will be a database or directory with has search options for finding and filtering projects. The third step will be to build a collaboration tool, things like Github integration, smart search and project management.

For now, Pfaff wants to see what the community response is and will make more developments from there.


VB's research team is studying mobile user acquisition... Chime in here, and we’ll share the results.