Text Mining-based Approach for Forecasting Spare Parts Demand of K-X Tanks

Abstract

One of the critical tasks of the defense logistics is the demand forecasting of spare parts, Because low-toned accuracy can lead to substantial budget wastes, Each military used the information management system to analyze the past spare parts consumption data information and predicted the demand of each part in a time series. However, a low-toned accuracy of the demand forecasting should be improved. In our study, we gathered a large amount of spare part consumption data first and derived several features including unstructured textual data to utilize them in the discrimination of fastidious patterns in the spare part consumption data. Our approach shows improved performance in demand forecasting with higher quantitative accuracy. The result shows better prediction accuracy than the existing time series.