Using graph visualization of call detail records to assist crime reconnaissance

• Main Authors: Hao-Wei He, 何浩瑋
• Other Authors: Sun, Tien-Lung
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/45023243166038550246

碩士 === 元智大學 === 工業工程與管理學系 === 92 === Call detail records contain lots of useful information for crime reconnaissance. Examples of information contained in call detail records include time patterns of activities, relationship of suspects, and accomplice. In the past, call detail records have been analyzed mainly by using statistical methods or directly examination. Statistical methods could not reveal details of the data; whereas direct examination is time-consuming and inefficient. In order to solve these problems, we bring up visualization to offer an efficient analysis of call detail records. Call detail records contains two types of data: calling phone numbers and calling time. The network graph is used to represent the connection of phones, and time-graph for communication sequences. The graph patterns showing in the network graph that have special meaning in crime reconnaissance are discussed. The time-graph is used to support further analysis of these crime-related graph patterns. We then discuss how the three types of activities in crime reconnaissance could be supported by visualization. Finally, we introduce two real-world cases to explain the details of the proposed visualization-based crime reconnaissance analysis.