Enhancing POI Data: LA & NYC Wrap-up
It’s been a busy few months at StreetCred, from attending the excellent Esri User Conference and mapping San Diego, to wrapping up our first multi-city, multi-partner data launch in both Los Angeles and New York City. We’ve wrapped up similar launches before, but things are different this time.
First, we explored a new business model in LA and NYC, where we imported partner data and worked with the StreetCred community to improve it in targeted ways, like adding photos, hours of operation, and phone numbers. We imported nearly 15,000 partner records which included only names and locations (and in some cases, business categories). Some records were thought to be high-quality data for businesses that were still open, others were expected to have closed. Still others were of unclear quality. We tracked the StreetCred community’s performance on this data and are excited to say that we enhanced or removed nearly half the records, which were scattered across large areas of both cities. This is great progress for a month-long contest, and we’re learning how to scale this on an ongoing basis.
Second, we’re excited to announce the hire of Chris Shughrue, StreetCred’s new data scientist. He’s rolling up his sleeves with the results and will share some of our findings here in the coming weeks. Longer term, Chris will help us to build a system that predicts POI coverage globally, which we’ll use to map the world and understand the quality of various datasets. We’re excited to get started on this!
Third, MapLA and MapNYC were the last time-bound contests StreetCred will run. We’ve learned a lot about how and where users like to map, how we can automate data validation, and how to provide community incentives to meet customer demands. We’ve learned so much that we’re changing the model, and will share more about that soon. Early hint: we will be leaning into the game dynamics that our early users have enjoyed and given feedback about, but extending it to provide richer map data about wider areas over longer time periods.
It’s going to be fun!