In 2013, I was part of the Standby Task Force Typhoon Yolanda (Haiyan) Activation. We worked on this activation from Thursday November 7th to Monday November 11th. Standby Task Force was activated by UN OCHA through the Digital Humanitarian Network. We utilized MicroMappers, a new crowdsourcing, social data processing tool that has recently been developed by Patrick Meier and QCRI. The final maps were generated as an ArcGIS MapStory from a live tweet feed off of the SBTF geolocated and categorized spreadsheet. In this series of blog posts, I will discuss some of the major takeaways that I gleamed from my experience with the Typhoon Yolanda activation.
Digital Crisis Response is Powered by Pure Collaboration
The digital work environment of crisis response is unparalleled. Nowhere else will you find so many talented professionals from so many different disciplines and organizations. They are all fervently working together to achieve the same thing. These are the shared response goals, absent of any motive for profit, notoriety, status, or other base intention. The people that are helping in crisis response genuinely care and want to do anything they can to help. It’s beautiful!
As part of my work for SBTF, I received a request from ShelterCluster to build a scraper to extract, from the Philippines National Statistics Coordination Board website, population data down to the Barangay level for the entire country of the Philippines. While I have built tools in Python before, my skill level is not such that I could accomplish this task in short order. When I Facebooked about my fumbling along with scraper building, my friend, Anass Bensrhir, jumped in to offer his expertise. He is in France. He must have stayed up all night building the scraper and extracting data to the exact specification of ShelterCluster needs. The final product was about 300 MB and was too large to be loaded or manipulated in Excel.
Due to the syntax of the scraper, there were some spare characters that needed to be parsed out of the dataset in order to clean it up.
Another Facebook friend of mine, Maksim Tsvetovat, cheerfully jumped in to write a 14-line Python code that cleaned the entire dataset, all in less than a half hour. To be honest, I was flabbergasted. It’s because of the programming prowess and the generosity of these talented and pricey data science consultants. Thanks so much to both of you!! (I am still in the process of making sure this data becomes accessible and openly available for future crisis response deployments in the Philippines. I will put more time into this once I get settled at my nomadic home base in Chiang Mai, Thailand)
But it wasn’t just this help from my personal friends… During the activation, I came across and collaborated with individuals from at least 10 different organizations. Although we each had our own mandate and mission, we all helped each other as much as we could. There was no squabbling over who owned what data or who got what credit. Because this is in crisis response, and to digital responders, it really doesn’t matter. What matters is that we do as much as we can, as fast as possible, to get information out that can help aid and relief organizations that are deploying on-the-ground to save lives.
