DETERMINANTS FOR GLOBAL CARGO ANALYSIS TOOLS

摘要

The purpose of Global TRADER (GT) is not only to gather and query supply-chain transactional data for facts but also to analyze that data for hidden knowledge for the purpose of useful and meaningful pattern prediction. The application of advanced analytics provides benefits beyond simple information retrieval from GT, including computer-aided detection of useful patterns and associations. Knowledge discovery, offering a breadth and depth of analysis unattainable by manual processes, involves three components: repository structures, analytical engines, and user tools and reports. For a large and complex domain like supply-chains, there are many stages to developing the most advanced analytic capabilities; however, significant benefits accrue as components are incrementally added. These benefits include detecting emerging patterns; identifying new patterns; fusing data; creating models that can learn and predict behavior; and identifying new features for future tools. The GT Analyst Toolset was designed to overcome a variety of constraints, including lack of third party data, partial data loads, non-cleansed data (non-disambiguation of parties, misspellings, transpositions, etc.), and varying levels of analyst experience and expertise. The end result was a set of analytical tools that are flexible, extensible, tunable, and able to support a wide range of analyst demands.