Using an Import to Build Community: The Seattle Case Study SotM 2013

Using an Import to Build Community: The Seattle Import Case Study

CUGOS

Clifford  Snow,  [email protected]  Jeff  Meyer,  [email protected]    

OpenStreetMap  Sea+le  

Seattle Import Overview

•  Motivation •  Early learning •  Import Design •  Import Progress •  Lessons Learned

h;p://gerryd90.deviantart.com/art/Space-­‐Needle-­‐Orange-­‐Paint-­‐303441533  

Our Goal: Make OSM’s Seattle Map Better Looking & More Useful

SotM US 2012: Portland SteveC Challenge

Go get some addresses!!!

Early Plan

1.  Get data (and verify it… : ) )

2.  Import it

3.  Profit!!!

wait a sec… not so fast…

Actual Early Steps

1. Found the data & cleared the license issues

2. Followed wiki import guidelines 3. Sent note to imports@ 4. Fielded lots of email

1.  Main list: challenging, terse 2.  Off list: friendly, supportive

5. Re-wrote / re-orged wiki import guidelines

6. Built extensive Seattle Import wiki page

7. Answered even more email

Weathering Import Critiques…

1.  Imports discourage community 1.  Matt Amos blog post -

http://www.asklater.com/matt/wordpress/2009/09/imports-and-the-community/index.html

2.  Blog is based on simulation, not real world observations

2.  Imports destroy data 1.  Maybe… but anything can… 2.  & is that always bad? 3.  & what if the original author doesn’t mind…

3.  Imports degrade the map

1.  OSM measures of map quality? 2.  Accuracy, Completeness

OSM Has Great Tools for Divvying Up Community Workloads

LINZ-­‐2-­‐OSM  

HOT  Tasking  Manager  

h;p://linz2osm.openstreetmap.org.nz/  

h;p://tasks.hotosm.org/  

Overall Plan

1.  Recruit volunteers

2.  Prep the data

3.  Divide up data by neighborhoods contributors know

4.  Curate the import in a very artisinal process

5.  Augment import with walk-arounds

Recruiting the Team

CUGOS

“Artisinal” Data Import 1.  Download import data for

region of choice 1. Prepared by Seattle OSM

team 2.  Load that data as layers in

JOSM 3.  Compare to Bing imagery 4.  Edit existing area in JOSM 5.  Cut and paste… very

carefully 6.  Add additional data, and

remove inaccurate data, as appropriate during import

7.  Add to data with walk-arounds

JOSM  at  the  desk…  

Go  Map!  in  the  Field  

h;ps://itunes.apple.com/WebObjects/MZStore.woa/wa/viewSoUware?id=592990211&mt=8  

Training – Driven by Internal Events

Import is a great excuse to get together. December 2013 Start 9 events / meetings in 6 months

4 Import-specific / JOSM training 5 local, walk-around focused events

Imports can Open Doors… Imports are a great excuse to barge into other meetings and recruit new mappers. ….and… imports can provide a great data set for training.

CUGOS

Import Sample

User: Chronomex Neighborhood: Capitol Hill

Adding Local Knowledge

User: Chronomex Neighborhood: Capitol Hill

Seattle Import by the Numbers

1.  6 months so far… 2.  9 events 3.  20+ Participants 4.  Need to verify # of

edits / updates 5.  Many mappers 6.  1 Reddit holla!

What might make this bigger, faster, more involved, more productive?

Next Steps

1. Finish the import 2. Double-check / review the data

1.  Review for errors 2.  Standardization

3. Walk-Arounds - focusing attention on: 1.  Neighborhoods 2.  Business data 3.  Parks

4. Outreach to civic partners 1.  Park micromapping 2.  Accessibility routing

Lessons Learned

1.  OSM import processes are unnecessarily onerous –  Biggest problem: matter of perspective - for community, for data –  Overall process, review steps – WG should help –  No authoritative source for guidance, decision making –  Separate accounts are hard to explain to new users and split a user’s identity –  Needs to be more n00b friendly

2.  All objections to imports are addressable –  Quality –  Community

3.  A big pile of data can draw mappers like bees to flowers –  More data can require more people, which can generate additional

exposure / publicity

4.  Import data can encourage field mapping