PNNL

Abstract

Transactions are fundamental components of an economy. This paper presents an analytic apparatus that can be used to analyze transaction data, where the transaction is the fundamental unit of observation. Transactions also are a potentially fundamental observation associated with the detection and characterization of organizational activities and events through acquisitions, trades or financial transactions. The objective of the research described in this paper was to develop a mathematical signature that represents transaction data (Point A to Point B, etc.), and visualize the transactions using currently available visualization tools. The representational signature should be useful for indicating change in organizational behavior, and for indicating when anomalous behavior occurs, i.e., something that is different than the common daily, quarterly or annual occurrence. The mathematical construct will be the same whether the transactions are country trade data or bank transactions, electrical grid transactions or some other multi-point transfer of information, asset, action, etc. The particular data example shown in this paper is international economic trade data for six countries; Mexico and its 5 largest trading partners, the United States, Germany, Canada, Japan and South Korea.