340 likes | 701 Views
Internet Traffic Search and Ethnic Relations in Russia. The Promise. Jeremy Ginsberg, Matthew H. Mohebbi, Rajan S. Patel, Lynnette Brammer, Mark S. Smolinski & Larry Brilliant, “Detecting influenza epidemics using search engine query data,” Nature Vol 457, 19 February 2009
E N D
The Promise Jeremy Ginsberg, Matthew H. Mohebbi, Rajan S. Patel, Lynnette Brammer, Mark S. Smolinski & Larry Brilliant, “Detecting influenza epidemics using search engine query data,” Nature Vol 457, 19 February 2009 • Developed a method to analyze a large volume of Google queries to detect social behavior • Improved early detection of social behavior http://dx.doi.org/10.1038/nature07634
Key components of the study • Google Trends: Internet search traffic measurement • Search query database: including IP addresses associated with each query • Prior years’ hard data on the behavior of interest (CDC) • Automated query-selection process: Identification of 45 highest-scoring search queries that fit the data from prior years • linear regression with 4-fold cross validation • Regional data aggregation: Maximizing fit probability and minimizing false positives • fit models to four 96-point subsets of the 128 points in each region. • Final model fitting: Tests in later years and for each individual state
Key lessons • Sensitivity to the nature of social behavior • Sensitivity to variation across regions • Sensitivity to variation over time • Multiple measures • Data reduction
Applications for Ethnic Conflict ResearchBuilding Block 1:Search Engines & Traffic Calculators
Applications for Ethnic Conflict ResearchBuilding Block 2:Search Query Sources
SOVA Information-Analytical Center Project: “The Language of Hate in the Mass Media”http://xeno.sova-center.ru/213716E/21398CB
Other sources • Internet buzz: chat forums, blogs • Tube traffic: cell phone clips, provocative song titles, video posts • Yandex search spin-offs (trends by words) • Focus groups • Expert groups
Applications for Ethnic Conflict ResearchBuilding Block 3:Behavioral Data Sources
Violence Data • SOVA Center (xenophobic violence; systemic interethnic communal violence) • UCSJ: Anti-Semitism • Кавказский узел (хроники по Ингушетии, Дагестану, Чечне) • Voinenet.com (weekly event summaries across the North Caucasus) • Data on protest attendance (police, HR NGOs) • GEDS archives (UMD)
Voting datahttp://www.vybory.izbirkom.ru/region/region/izbirkom?action=show&root=1&tvd=100100021960186&vrn=100100021960181®ion=0&global=1&sub_region=0&prver=0&pronetvd=null&vibid=100100021960186&type=233
Potentially trackable phenomena • Mobilization: • Xenophobic group names; party names; leader names; event names • Intergroup hostility • Expressions of hate (derogatory group epithets) • Violence: • Proxies (weapons purchase, martial arts clubs, extremist forums names) • Minority groups’ resistance: • Specific ethnic group names + rights; HR NGOs; defense lawyers names, etc • Trust in government/state capacity: • Legal texts, statutes (Art. 282), police measures