1. Introduction
1.6. Literature Review
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
1.6. Literature Review
1.6.1. Export Controls in Public and Private Sectors
Currently, the major methodologies adopted by countries, such as US and Germany etc., are focusing on ICP (Internal Compliance Program) and end-user screening.
In the public sector, the ExportCont.org introduces in the Overview of the U.S. Export Control System that preventive enforcement should include established procedures related to export license applications (i.e. screening the proposed item, quantity, end-use and all parties involved in the transaction for any potential export) and compliance mechanisms (i.e.
working in partnership with industry to educate them on how and why to monitor and control their own export activity).6
Georg Pietsch, Director General of Germany Federal Office of Economics and Export Controls, emphasizes that the end-user screening of export control should ensure a differentiated approach, especially with regard to:
be able to properly evaluate each individual case,
the different sensitivity levels of consignee states concerned,
the reliability of the persons involved in the transactions,
and the principle that sensitive cases need more efforts than cases of less sensitivity.7
In addition, Pietsch explains the criteria of plausible licensing check are:
technical aspects,
internal knowledge and other information of the authority,
submitted documents and papers,
6 ExportControl.org, "Overview of the U.S. Export Control System," ExportControl.org, http://exportcontrol.org/links/2081c.aspx (accessed Nov, 12, 2011).
7 Georg Pietsch, "End use End-User-Screening" (Seventh Annual International Export Control Conference, Stockholm, Federal Office of Economics and Export Control (BAFA) Germany, 20-22 Sep, 2005, 2005)., 1.
‧
reliability of the persons involved in the transaction (exporter, consignee, end-user and other persons involved),
risk analysis.8
Pietsch further elaborates that in German the backbone of risk analysis is a database with information about applicants, countries of concern, purchase, consignee, end-user and military or civil projects. The purpose of the end-user screening is to support decisions, to discover front-or mailbox-companies, procurement networks and other sensitive projects, overall to minimize the risk to diversion to proliferation end-use.9
In the private sector, there are compliance programs developed by IT or consultant companies. For example, the eCustoms–MSR Inc. provides Visual Compliance to assist customers to conduct comprehensive denied party screening for export shipments. Visual Compliance provides documentation and on-line tools to help avoiding non-compliance with U.S. Export Control Regulations. The Visual Compliance software solutions system performs Restricted Party Screening against all relevant U.S. Government lists, including Specially Designated Nationals and Blocked Persons (SDN).10 NextLabs, IBM, and SAP AG also developed an information export control solution to extend the export compliance capability of the SAP GRC Global Trade Services application. The solution manages the handling and export of technical information that is subject to International Traffic in Arms Regulations (ITAR), Export Administration Regulations (EAR), or other U.S. federal regulations.11
Basically, current export-control screening methodologies both in the public and private sectors are focusing on screening the entities which are already on the black lists. For the entities which are not on the black lists but used by the proliferators as conduits, governments usually rely on internal knowledge and other information of the authority.
However, the internal knowledge is not likely to cover all the applications since there are usually thousands of applications in a year. In addition, it is difficult to discover implicit
8 Ibid, 3.
9 Ibid, 6.
10 eCustoms, "Visual Compliance – Export Control Compliance, Comprehensive Denied Party Screening, Controlled Goods Classification, Export License Determination and U.S. Munitions List Classification for Your Export Shipments," eCustoms – MSR Inc.,
http://www.ecustoms.com/vc/vc.cfm (accessed Nov, 15, 2011).
11 Lawrence Dietz, "IBM, SAP, and NextLabs Collaborative Effort Yields Export Control Solution,"
Sageza Group, Inc., IT Analysis Communications Ltd.,
http://www.it-director.com/business/content.php?cid=9482 (accessed Nov 15, 2011).
‧
technologies to develop a methodology to discover implicit signs indicating potential violations to improve and supplement the end-user screening process.1.6.2. Categories of Data Mining Technologies
Richard J. Roiger and Michael W. Greatz define data mining as a process of acquiring knowledge by using computer technology to analyze information from databases. They note that the goal of data mining is to discover trend and patterns from data.12
For using data mining tools to solve real-world problems, Christopher Westphal and Teresa Blaxton noted in Data Mining Solutions, “When you begin a data mining engagement you have to make a series of decisions concerning such issues as how to integrate data from separate sources, what types of models to implement, what display paradigms to use, and which tools to choose for the application environment ―.13 There are a variety of commercial tools for data mining application. How can we choose appropriate data mining tools for handling export control issues?
Christopher Westphal and Teresa Blaxton introduce data mining systems in three paradigms of link analysis, landscape display, and quantitative analysis. They provide descriptions of some link analysis tools including NETMAP, Analyst‘s Notebook, Imagix 4D, Daisy, and others. They also discuss several tools that use landscape visualization, including Mineset 2.0, Metaphor Mixer, In3D, and others. Regarding quantitative analysis, they introduce tools of Clementine, Enterprise Miner, Diamond, and CrossGraphs. In addition, they also provide discussion of future trend in data mining technology, such as web visualizer, free text visualizers, full scope systems.14
In another way, Shia Ben-Chang categorizes data mining tools into case-based reasoning, data visualization, fuzzy query and analysis, knowledge discovery, and
12 Richard J. Roiger and Michael W. Geatz, Data Mining: A Tutorial-Based Primer, trans. 曾新穆 and 李建億, 初版四刷 ed. (台北市: 台灣培生教育出版股份有限公司, 2006), 393, 4.
13 Christopher Westphal and Teresa Blaxton, Data Mining Solutions: Methods and Tools for Solving Real-World Problems, Vol. 1 (USA: Wiley Computer Publishing, 1998), 617, 441.
14 Ibid, 198.
‧
neural networks.15 Data mining also can be categorized into six fields, including database systems, machine learning, statistical and data analysis methods, visualization, mathematical programming, and high performance computing. The major functions of data mining include classification, estimation, prediction, affinity grouping, and clustering.16
There are a wide variety of applications by using data mining, including pharmaceutical research, telecommunications, banking, retail sales and marketing, stock market analysis, and money laundering investigations.17 Data mining can also be used in customer profiling to predict potential customers, understanding customer behavior to decide logistic flow, or customer relations to maintain current customers. Recently, telephone, credit card, insurance, and government agencies are highly interested in fraud detection for decreasing loss. In addition, data mining can help us on detecting signals of customers with bad credit records, and we can use the signals to screen customers with high risks.18
1.6.3. Gaps of Current Methods
Some of the current applications have similarities to the management in export control, for example, money laundering investigation, fraud detection, screening bad-credit customers, and criminal investigation. Those applications are focusing on detecting problematic individuals from a whole or following leads to find the targets.
For developing the methodologies to deal with export control issues, there are some research puzzles that we need to solve as follows.
15 謝邦昌 and others, 資料採礦與商業智慧SQL Server 2005: Data Mining & Business Intelligence with SQL Server 2005 (Taiwan Taipei: 中華資料採礦協會, 2005), 381.
16韋端 and others, Data Mining概述: 以Clementine 7.0為例, 初版一刷, 2003), 885.Ibid.
17 Westphal and Blaxton, Data Mining Solutions: Methods and Tools for Solving Real-World Problems, 617, 442.
18 韋 and others, Data Mining概述: 以Clementine 7.0為例, 885.
‧
What is the essence and logic of the problems in export control?
What are the difference between the application in export control and other fields?
Besides of statistical methods, if using link analysis, which is used widely in investigation, is a proper way to build an effective data mining solution for export control?
By evaluating current application of data mining tools in various fields, the thesis will try to integrate investigation logic into data mining methodologies to develop proper solutions for improving export control management. The research will hopefully fulfill the gaps which have not yet been addressed in previous research, including using the data mining applications in export control, and enhance the research in using investigation method in data mining solutions.