CLIMBER RISK PROFILE MAPPING AND CLUSTER ANALYSIS OF CLIMBING ROUTES: A CASE STUDY OF MOUNT SINDORO, TEMANGGUNG, CENTRAL JAVA

Abstract

Hiking is a popular form of tourism. However, the number of injuries and casualties resulting from inadequate preparation is a significant concern. Previous research has focused on mapping visualisation, calorie measurement, and climbing logistics on hiking trails, and on developing hiking simulations based on climate and climbing capacity. Furthermore, risk mapping on hiking trails and risk assessment models are needed. This still requires further study, given that the difficulty level of hiking in mountainous areas in Indonesia lacks clear standards. Therefore, it is necessary to develop trail risk mapping that links climber/porter factors to environmental conditions to provide recommended risk assessments for specific profile conditions. This can be used to evaluate existing models and assess the risk of hiking trails in general, as well as specifically for the Mount Sindoro Trail in Temanggung, Central Java. Two approaches were used. The results showed that the SVM algorithm was able to classify the risk profiles of climbers with a Support Vector Machine (SVM) accuracy of 90.35%. Meanwhile, clustering using the K-Prototype method supports optimisation of trail management, facility efficiency, and decision-making for risk mitigation. This research requires further validation to provide data-based recommendations to relevant agencies to support improvements to the hiking trail information system in Indonesia.