Synthetic data creation for forensic tool testing: Improving performance of the 3LSPG framework

Increasing amounts of data require improvements in effectiveness and efficiency of forensic tools. If new tools have been developed, they have to be evaluated, e.g. by applying test data. 3LSPG has recently been proposed as a framework for generating synthetic test data by simulating activities of subjects using Markov chains. However, the generation of test data should also be efficient. In this paper, we show how to improve the efficiency of 3LSPG considerably compared to its original proposal. We show how to speed-up the calculation of state transition probabilities in the Markov model of 3LSPG by proposing an algorithm that is much faster and more reliable than the one originally used. The simplex algorithm serves as basis for our algorithm although it is typically used for the different purpose of solving optimization problems. Our algorithm helps to enable the creation of synthetic data for forensic tool testing with 3LSPG in significantly shorter time.