An Overview of Data Mining and Process Mining Applications in Underground Mining

Abstract

The underground mining process can be analysed with a data-oriented or process-oriented approach. The first of them is popular
and wide known as data mining while the second is still not often used in the conditions of the mining companies. The aim of this
paper is an overview of data mining and process mining applications in an underground mining domain and an investigation of
the most popular analytic techniques used in the defined analytic perspectives (“Diagnostics and machinery”, “Geomechanics”,
“Hazards”, “Mine planning and safety”). In the paper two research questions are formulated: RQ1: What are the most popular
data mining/process mining tasks in the analysis of the underground mining process? and RQ2: What are the most popular data
mining/process mining techniques applied in the multi-perspective analysis of the underground mining process? In the paper sixty-
-two published articles regarding to data mining tasks and analytic techniques in the mentioned domain have been analysed. The
results show that predominatingly predictive tasks were formulated with regard to the analysed phenomena, with strong overrepresentation
of classification task. The most frequent data mining algorithms is comprised of the following: artificial neural networks,
decision trees, rule induction and regression. Only a few applications of process mining in analysis of the underground mining
process have been found – they were briefly described in the paper.