In wjones127/thesis: Bike Routes Thesis (Title Case)

Week 2

This week, I met with William Henderson and Evan Heidtmann at Ride Report to talk about their data and read over the paper on ambulance data from last week.

Meeting with William

William is very supportive of me working with the data. Their goal with their project is to release an API that orgnizations will want to pay for later, and they are allowing researchers to use their data so that he can get an idea of what kind of API endpoints should be built in at the beginning.

One thing they need to balance is getting the right data to people while protecting the privacy of people who contribute to the data. Thus, if I want to run models on data from individual rides, I will have to bring in a script to run and have them run it on the data, rather than play with the data. They are very willing to do this.

In the meantime I do have that sample from Rennie's project. It is not matched to road though. They offered to give me a new dataset from some consenting bicylcists for that, or I could snap the data to the roads myself. I might consider doing that myself.

They had mentioned that other people are working on different angles of the data, including using the recorded routes to infer bicyclists preference of roads. It is probably notable that missing data in this case has meaning: a roads through which no one rides is likely not a good road to ride on, even if there are no ratings.