SARS risk profile indicator model evaluation

Abstract

Managing trade efficiently and in a cost-effective manner can greatly benefit the economy of a country. In South Africa, customs processes lead to many unnecessary delays seemingly without any underlaying reason. This research report aims to show how this conclusion was made, along with a range of other statements about the efficiency and state of customs processing in this country. The first chapter gives some background information, as well as states the main problem to be addressed, namely developing an application that can analyse raw customs data provided by SAAFF in order to reach meaningful conclusions from the data. The objectives for the project as well as the methodology are discussed in detail. Furthermore, a literature study showing motivation for the research and background on the more important topics are provided. The conceptual design broadly explains the design phase for the complete project, including all of the functional units and a complete functional analysis of all the units and the interfaces. The detail design then zooms in on the unit responsible for creating the analyses and provides flowcharts to explain the design of the analyses. The next chapter provides results after the implementation of the design and evaluates the results in detail. There are two main categories for the analyses, namely the statistical and the modelling categories. The former uses descriptive statistics to process the data in a meaningful way and to ensure easy interpretation for users of the software. The results are displayed in the form of histograms. The second category, modelling, makes use of statistical principles to deliver ruling on the probability of future occurrences. This could be of great help to SAAFF members when making decisions about their trades and can help them make more informed decisions in regards to the commodities they trade. The last chapter draws conclusions about the results and gives further recommendations on how the project can be improved upon. The attached appendices give more detail about the statistical tables used as well as descriptions for some of the customs codes used in the analyses. The user manual for the software is also included, along with the source code and full digital archive.