Log in / Register

Home arrow Computer Science arrow Social Computing, Behavioral-Cultural Modeling and Prediction
< Prev   CONTENTS   Next >

2 Related Work

A number of studies ([1], [2], [3]) have accounted for the qualitative statistics of racial segregation in major U.S. metropolitans, and how its extent has changed over time. Most of these studies use one or more of the five indexes explained by Massey and Denton [4] to compare the magnitude of segregation between two racial groups. Analyzing data from the 1980 census in 60 U.S. metropolitans, Denton and Massey's [1] findings indicated that blacks were highly segregated from white in all socioeconomic levels, relative to Hispanics or Asians. Although the levels of segregation has declined modestly during the 1980s [5], through the 1990s [2], and up to 2000 [3], blacks still remain more residentially segregated than Hispanics and Asians. Clark [6] points out that although a certain degree of racial integration is acceptable it is unrealistic to expect large levels of integration across neighborhoods, because there exists a tendency for households of a given race to cluster with others of similar race [7].

Veering from the conventional studies that measure the extent of segregation, few researchers have tried to identify social problems arising as a result of it. Peterson and Krivo [8] studied the effect of racial segregation on violent crime. Card and Rothstein [9] found that black-white SAT test score gaps during 19982001 were much higher in more segregated cities compared to nearly integrated cities. In our study, we consider another interesting effectbiases in mobility patterns as a result of segregation, and at the same time shed some light on the extent of segregation in the 2010 census data.

Spatiotemporal models of human mobility have been studied on various datasets, such as circulation of US bank notes [10] and cell phone logs [11]. Temporal human activities like replying to emails, placing phone calls, etc. are known to occur in rapid successions of short duration followed by long inactive separations, resembling a Pareto distribution. The truncated power law distribution characterizing heavy-tailed behavior for both distance and time duration of hops between subsequent events in a trajectory of normal travel pattern has been established by many studies [10], [11]. Although geo-location based data from Twitter has been used in several applications like spotting and tracking earthquake shakes [12] and street-gang behavior [13], it has not been used to model effects of racial segregation on mobility patterns and behavior.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science