The organizers of CrisisCamp Sydney exceeded their expectations and objectives for the geocoding and crowdsourcing event planned for the weekend despite a rainy start on Saturday. Fifteen volunteers joined the first day and nine showed up on Sunday. We were lucky to have very talented, hardworking and committed volunteers. Here's a wrap up of what happened.
Originally, we were hoping to group the number of volunteers depending on their skill sets. In every group, we planned on putting a developer or two, a researcher and one or two members with Pakistani language proficiency. But as everyone got more settled with geocoding, a number of our developer experts were able to improve on the work flow specifically in making the geocoding less laborious. In just a few hours, they were able to reconfigure how to make the workload more efficient.
Paul helped us in designing a more efficient work flow by creating GIS layers which added context to the mashup application. From this, Luke wrote a script that summarized village data on the OSM. Meanwhile, we had our other volunteers, Adam, Fadhillah, Umair, Eva, Joel, Warren, Rene and Ping continue to do geocoding and crowdsourcing (using Crowdflower). Another volunteer, Lachlan, was tasked with helping automate some of the tasks needed to be performed on Crowdflower. These allowed the tasks to be automated and made it easier for our geocoding and crowdsourcing volunteers. He did this using a grease monkey script. One Pakistani volunteer, Anita, was also connected with Anihi at CrisisCommons to help with translating in Urdu a wiki page.
Another important task completed was when Pamela, Hassan and Ivan, using a JavaScript/Mashup, created an application whereby the villages that had been geocoded were in bold and this information was then matched against the spreadsheet of flood affected villages found a http://bit.ly/pkflood_kp.
As the first day rolled along, our volunteer videographer, Tolmie shot and produced a tutorial video on geocoding villages using OSM and US DMA data. After editing, he uploaded it on Youtube and the CrisisCommons wiki. It will also be uploaded in the Drumbeat site.
On both days, Aram, wrote an application using Django that pulled out of Crowdflower messages based on their urgency, prioritizing these messages to be processed by crowdsourcing volunteers.
All in all, our team had processed over 200 SMS messages in Crowdflower. We added about 300 or so villages on the Open Street Maps.
Our student partners, Khiem and Sandy have done a great job in helping us secure the venue, helped with the catering and the other logistical details that needed to be attended to for the two day event. Martin has shown his support, sponsorship and troubleshooting when necessary. Organizers Shoaib and Vicky, after doing all the coordination, finding partners and planned how the event would unfold, were running around doing all sorts of tasks that needed to be done druing the two days, aside from providing instructions and orientation on geocoding, crowdsourcing. In between, the two were also seen geocoding, crowdsourcing, writing blogs, monitoring all the channels with other CrisisCamp groups in Toronto, Bangkok and London, making contact on the IRC web chat node, skype, etc. We thank all our sponsors and partners: CrisisCommons, CIE, Humanitarain OpenStreetMaps, Mozilla Drumbeat, SIFEUNSW, and the World Bank.
To all our volunteers - what a fantastic group of people. You've done us all proud!!!! Big, big thanks.
