Abstract—A co-offending network is a network of offenders who have committed crimes together. Recently different researches have shown that there is a fairly strong concept of network among offenders. Analyzing these networks can help law enforcement agencies in designing more effective strategies for crime prevention and reduction. One of the important tasks in co-offending network analysis is central actors identification. In this paper, firstly we introduce a data model, called unified crime data level and co-offending network mining level. Using this data model, we extract the co-offending network of five years real-world crime data. Then we apply different variations of centrality methods on the extracted network and discuss how key player identification and removal can help law enforcement agencies in policy making for crime reduction. I.

Abstract We study peer effects in education. We first develop a network model that predicts a relationship between own education and peers' education as measured by direct links in the social network. We then test this relationship using the four waves of the AddHealth data, looking at the impact of school friends nominated in the first wave in 1994-1995 on own educational outcome reported in the fourth wave in 2007-2008. We find that there are strong and persistent peer effects in education since a standard deviation increase in peers' education attainment translates into roughly a 10 percent increase of a standard deviation in the individual's education attainment (roughly 3.5 more months of education). We also find that peer effects are in fact significant only for adolescents who were friends in grades 10-12 but not for those who were friends in grades 7-9. This might indicate that social norms are important in educational choice since the individual's choice of college seems to be influenced by that of friends in the two last years of high school.

Abstract This paper illustrates a procedure to estimate externalities from indirect connections (so-called network externalities) using network data. This structural approach is suitable for static games of endogenous network formation with partial information, and relies on the equilibrium conditions of pairwise stability

Empirical analysis of social networks is often based on self-reported links from survey data. How we interpret such data is crucial for drawing correct inference on network effects. We propose a method to test whether survey responses can safely be interpreted as a link and, if so, whether links are generated by a unilat-eral or bilateral link formation process. We present two empirical illustrations of the test on risk-sharing links in Tanzania and on communication among Indian farmers respectively, demonstrating the ability of the methodology to discrimi-nate between competing data generating processes.

The formation of migrant networks

(ii) development of