Linking and Visualizing Social Media Data about Crises

Abstract

Making sense out of data is a crucial process in decision making. This is especially so when it is for saving lives in crises and disasters like earthquakes, oodings or forest res. Social media is a prominent source for human observations about disasters. Linking these crowdsourced observations together with SMS messages sent to aid phone numbers carry potential information about not only buldings collapsed or roads blocked, but also about people in need for medical services, food and water. This calls for linking the di erent pieces of information together by making temporal, spatial and thematic references explicit and groupable. We argue that this supports information usability and would thus create grounds for decision making. Examples of social media data and its handling about the earthquake in Haiti on January 12, 2010 serve in this to illustrate both the research problems and solutions.

-

terms originate from various state-of-the-art sources including the Inter Agency Standing Committee (IASC), Emergency Shelter Cluster in Haiti, UNOCHA 3W Who What Where Contact Database and the Ushahidi platform.

Haiti earthquake being the rst disaster where large scale crowdsourced information was gathered [ 3 ] serves as a case for understanding challenges one has about data. For this Linked Haiti is a dataset available and accessible online3 for exploring and understanding the Haiti Earthquake in 2010. For instance, Figure 1 illustrates the power of grouping social media messages by categories and visualizing the results as interactive bubbles4. 3 http://linkedscience.org/data/linked-haiti/ 4 This technique is realized by the help of the D3.js library

References

1. Minu Limbu . Integration of crowdsourced information with traditional crisis and disaster management information using Linked Data . Master's thesis , University of Muenster, Germany, 2012 .

2. Jens Ortmann , Minu Limbu, Dong Wang , and Tomi Kauppinen . Crowdsourcing Linked Open Data for disaster management . In Rolf Grutter, Dave Kolas, Manolis Koubarakis, and Dieter Pfoser, editors, Proceedings of the Terra Cognita Workshop on Foundations , Technologies and Applications of the Geospatial Web , In conjunction with the International Semantic Web Conference (ISWC2011) , volume 798 , pages 11 { 22 , Bonn , Germany, October 2011 . CEUR Workshop Proceedings.

3. Matthew Zook , Mark Graham , Taylor Shelton , and Sean Gorman . Volunteered Geographic Information and Crowdsourcing Disaster Relief: A Case Study of the Haitian Earthquake. World Medical & Health Policy , 2 ( 2 ):7{ 33 , 2010 .