Intel ISEF 2013 Finalist Profile

CS069 (Park)
Where and When: Hometown Incident Analysis and Situational Awareness
Sang Jun Park
Homestead High School, Fort Wayne, IN

Rapid collection of crime reports have grown analytical needs of law enforcement. Criminologists are increasingly aware of the importance of places of crime because not only are places logically required, but also their characteristics influence the likelihood of a crime. Current crime analysis, however, uses macro approaches with looking at aggregates of incidents by regions and communities, rather than examining the places themselves. This study focuses on the micro-level examination of crimes, and presents a computational framework for discovering interesting relationships among crime places, nearby facilities, and their features. The framework has four major steps: 1) preprocess spatial and non-spatial (including temporal) attributes, 2) compute spatial relationships, 3) discover spatial association rules, and 4) finally evaluate the patterns using visualization and domain knowledge. The first step includes geocoding, reverse geocoding, discretization, and extraction. The second and third steps are time consuming tasks. Spatial indices and spatial buffering and topological relation are used for efficiently searching neighboring objects on space. An association rule mining algorithm is used for discovering spatial association patterns with high confidence. The final step includes the visualization of data points of result patterns on map. For the validation of the proposed approach, this project conducted the empirical study with local crime incident data and Points of Interest for the facility data. This study result ultimately helps law enforcement agencies to respond in the most effective manner. Furthermore, the proposed framework is applicable for analysis of other incident type data such as disease cases.

Connect with SSP
  LinkedIn  Facebook  Twitter