GeoWorld

GeoWorld February 2012

Issue link: https://read.dmtmag.com/i/55888

Contents of this Issue

Navigation

Page 12 of 31

Workshop and Conference held in Sapporo, Japan, Jan. 12-18, 2012, are working to make such tools fast and effective. Some software companies such as Google have developed specialized tools to data mine user searches. Google Trends and Google Flu Trends provide data on keywords used in searches and estimates of flu severity. Five of the 70 databases, including these two, are text-mining applications. For text mining and related applications to be useful, you will need high spatial and temporal resolution. The latter will be easier to produce reliably than the former. Searches for data patterns will be far easier to generate than raw socio-economic indicators. For example, if interested in prenatal sex selection, WACKE will give you sex ratios by country, but it won't generate lists of such data with high spatial and temporal resolution. Five of the databases, including LinkedData, are specifically designed for data sharing. These are really useful, and, over the 10-year life of the project, you can expect enormous advances in this area and in "query analytics." Eleven databases, the largest number, relate to conflict around the world. These are likely to be extremely useful and will impact economic forecasts and socio-economic development and many other modeling activities. The European base is likely to minimize North American centrism, but be sure to include non-European newsfeeds, such as Al Jazeera, among other world views. Please be aware that conflict isn't limited to declared and undeclared wars. Broaden your definition and add such data- bases as Ushahidi and FrontlineSMS to your list. Apart from the United Nations Conference on Trade and Development database, the six others listed relating to economics and finance are completely North American and/or Euro-centric. The world has changed and, by the end of the project in 2022, China will have the world's largest economy. Can you really ignore it? The Future ICT project proposes to use five scientific databases that list contributions to scientific knowledge. You are fortunate that nations such as China and India have succumbed to the domination of listing scientific knowledge by such organizations as the ISI Web of Knowledge and Google Scholar. Your North American/Euro-centrism is less of a problem here. The Social Sciences, Urban Data and Traffic Data categories include seven, two and nine databases, respectively. Is traffic quite so important? Again, you're showing too much bias to the United States and Europe. Did you know that China now is the No. 1 producer of cars in the world? And it has some of the worst traffic problems. Only two of your databases, the World Value Survey and the Global Urban Observatory Database are truly global. This is woefully inadequate. Have you heard of globalization? Three databases purport to deal with what Helbing calls "Open Maps." Here, for the first time, the databases are truly, explicitly spatial. Did you know that spatial scientists estimate that 80 percent of human activity takes place somewhere? Geography matters. And only the OpenStreetMap database is truly open, and only it exploits volunteered geographic information. Kudos to you for including Tracksource Brasil. This is your first database from South America. We are still waiting for one from Africa. There are two logistics databases (both U.S.- based), two climate databases and five open-data initiatives that primarily are U.S.- and UK-government based. The 70 databases are rounded out by a sole contribution in the "Reality Mining" category that includes nine months of mobile-phone data for 100 MIT students during the 2004-2005 academic year. Are you serious? Really? 100 MIT students—what sort of sample size is that? Please do not give away €100 Billion of research money until the Future ICT proposal does the following: 1. Recognizes the importance of spatial data and includes the large amounts of GIS data available from such companies as Esri and all national census organizations (see Hansen et al., 2011, for a discussion of Latin American data for election analysis). 2. Includes the vast repositories of freely available remotely sensed data. 3. Produces a better geographical coverage across the world and recognizes that there are countries outside of Europe and North America. 4. Discusses the importance of adequate spatial and temporal resolution as well as the need to develop appropriate data-mining procedures. 5. Recognizes the importance of social networking and communications such as the Twitter feed and cell-phone data. Reference Hansen, C.V., Sun, H. E. and Waters, N.M. 2011. "The Media Communications Environment through a Spatial Lens: The Mapping the Media in the Americas Project," Canadian Journal of Communication, Vol. 36, No. 1, pages 141-145. FEBRUAR Y 2O12 / WWW . GEOPLA CE . COM 13

Articles in this issue

Archives of this issue

view archives of GeoWorld - GeoWorld February 2012