Background The global rate of aviation accidents has recently been stabilising and the situation can now be regarded as satisfactory, but because of the growth in air traffic, the absolute number of fatal accidents per year might increase, if the flight safety will not be improved. The collection of data and reporting systems have reached their top level. The focal point in increasing the flight safety is analysis.
Methods The source of aviation safety data are those from the aviation field collected deviation and incident reports that include both structured and narrative fields. 1200 flight safety reports from a three-year period were used as test material. The narratives of these written in Finnish were processed with three text mining tools applying clustering. One is totally language independent, the other has a specific configuration for Finnish and the third was originally created for English, but encouraging results achieved with other languages, a Finnish test was undertaken, too. The totally language independent one is a Finnish prototype created in one of the Universities of Technology, the two others commercial products. The mining was carried out by performing one round with all the systems and the second with two of them in order to get more accurate mining results after refining the mining definitions.
Results It is obvious that in case events leading to lethal trends would have existed in the data, they would have been discovered and brought out. The text mining tools used were capable of extracting trends – actually recurring events – that turned out to be incidents. However, in the cases studied they did not develop into dangerous risks or accidents.
Conclusions All systems provided encouraging results, as well as proved challenges still to be won. Flight safety can be significantly improved through the development of data analysis. Narrative text mining is demanding also because of the multiplicity of languages spoken in the world.
- Aviation safety
- lethal trend
- text mining