Using GUHA Data Mining Method in Analyzing Road Traffic Accidents Occurred in the Years 2004–2008 in Finland

Abstract

The suitability of the GUHA data mining method in analyzing a big data matrix is studied in this report in general, and, in particular, a data matrix containing more than 80,000 road traffic accidents occurred in Finland in 2004–2008 is examined by LISp-Miner, a software implementation of GUHA. The general principles of GUHA are first outlined, and then, the road accident data is analyzed. As a result, more than 10,000 associations and dependencies, called hypothesis in the GUHA language, were found; some easily understandable of them are presented here. Our conclusion is that the GUHA method is useful, in particular when one wants to explore relatively small size, but still significant dependencies in a given large data matrix.