地球・資源分野ショート講演

[2K0101-12-11] Rating System Creation and Dataset Attributes Similarity Analysis from Sedar Database Website for the Development of Mining Method Recommender System

○Elsa Pansilvania Manjate¹, Youhei Kawamura¹, Adachi Tsuyoshi¹, Jae won Kim¹, Bernardo Bene² (1. Akita University, 2. High Polytechnique Institute of Tete, Mozambique)

Keywords:Rating System, recommender system, decision-making, mining method

Mining method selection is one of most critical and complex decision-making task in mine planning due to the need to consider several factors (technical, economic, political, social and historic), therefore, developing a recommender system would highly contribute for improving the efficiency and accuracy of mining method selection task. The purpose of the research is to develop a mining method recommender system based on mining companies’ technical reports downloaded from a Canadian Sedar Website. So, for mining method selection are considered orebody characteristics (physical and mechanical) and mining-technical and economic factors(parameters). The initial steps of the research were to get data and then analyse into document management software, in which, a searching system was created in order to simplify data quality and quantity analysis. Therefrom, analysed data are used as a database for the creation of input data matrix for recommender system development, where the main elements of the matrix are mining method selection parameters got from each mining company technical report. In addition, an appropriate rating system and placing sequence of each parameter are defined by measuring the impact that each parameter or attribute has in mining method selection.