A data mining approach to analysing airborne wood particulate concentration and atmospheric data

摘要

Exposure to airborne wood (hard and soft) dust can lead to a number of diseases, such as asthma, emphysema, bronchitis and upper respiratory tract cancers, lately even proven to be linked to elevated risks for chromosomal instability in cells of the aerodigestive tract. In this context, the paper investigated the particulate wood dust concentrations in a university environment near a timber mill using a data mining approach consisting of JRip, J48 algorithms and a multilayer perceptron (MLP). The data collected consists of particulate wood concentrations and related atmospheric conditions recorded over a few days at four different locations within the university situated next to the timber mill. The results reveal that ORICC is the location most exposed to high concentrations of wood dust (up to 1.57 MG/M at times). This exceeds the recommended exposure limit of 1 MG/M for humans if the dust particles were of hardwood hence, more tests are recommended to establish the airborne particulate wood dust composition from the factory.