OUTLAW: Using Geo-Spatial Associations for Outlier Detection and Visual Analysisof Cargo Routes

U.S. Customs deals with a huge number of cargo trucks and shipments crossing borders by air, water and land. In this paper, we present a system for Outlier analysis by measuring waywardness, called OUTLAW. It distinguishes abnormal and wayward behavior of cargo or goods from that of normal. This wayward behavior can appear to be normal due to lack of correlation. Our aim will be to combine disparate data in meaningful ways by utilizing such parameters as spatial proximity, spatial correlation, and association. Use of thematic map coloring, geographic visualization of individual variables can be very effective for identifying correlations between the variables, week spots, loop holes, wayward routes or vagrants etc. Based on the correlation of the data a predictive model can be generated to detect an index of measuring the waywardness. OUTLAW employs an N-step mechanism to detect vagrants or outliers in the normal scheme of events.