[talk-ph] armchair mapping Quiapo area

[maning sambale]

Dear OSM-PH and Hotties,

Just a word of thanks to those who helped in mapping Quiapo area [0].
Your contribution helped a lot in the buildup of our dataset for our
DRR project.

After our mapping event, we believe we covered 50% of all buildings in
the area [1].  A subsequent field mapping was done by our staff and
local partner which increased the coverage to ~70%. We will continue
validating the data and in the coming days, test the suitability of
the information into InaSafe's impact modelling [2].

There are a few more tasks that need to be completed in the extended
areas around Quiapo, and, we encourage you fill in the gaps.

One thing we observed during the validation is that while the initial
tracing is very helpful, sometimes, "poor" tracing can influence the
quality of validation in the field using field/walkingpapers
particularly for highly urbanized areas like our site.  Some research
calls this path dependency effect, where, “the 
first iterations have
a high impact on the final results due to a path dependency  effect.”
A good discussion I found was from Patrick Meier's blog [3].


[0] http://tasks.hotosm.org/job/56
[1] http://www.openstreetmap.org/?relation=2524687
[2] http://inasafe.org/
[3] http://irevolution.net/2012/09/24/best-way-to-crowdsource-satellite-imagery-analysis/