OpenStreetMap and Hadoop

Question

I need ideas for an exit project on Hadoop and OpenStreetMap.

I have access to an AWS EC2 instance with an OpenStreetMap snapshot in my EBS volume. OpenStreetMap data is in the PostgreSQL database.

What MapReduce function can I run in OpenStreetMap data if I can export it to xml format and then put it in HDFS?

In other words, I now have a brain spasm, and I can’t imagine what MapReduce operation can be learned from understanding OpenStreetMap xml?

(i.e. retrieve all the places designated as a park or golf course, but this must be done once, but not constantly)

Many thanks

Answer 1

• An interesting idea of ​​my colleagues is to find nodes that are not connected, but should be. For example, a street that suddenly ends and continues for another 1 m along a separate path.
• Pre-calculate the shortest paths