Protected Wildlife Identification System using Decision Tree based on Morphological Features to Support Wildlife Conservation in Indonesia

Abstract

Abstract Most endemic wildlife in Indonesia is under serious threat, as their habitat is lost due to forest clearing and illegal hunting by trade associations. A significant challenge that hinders law enforcement of illegal wildlife trading is the lack of knowledge and inconvenient methods for identifying the wildlife (protected or not). We suggest an identification system by implementing mobile and web applications. A decision tree is implemented in the application to identify the wildlife. Seven hundred thirty-two species listed on the Minister of Environment and Forestry Regulation No. 106 of 2018 are used in this study. The decision tree for each group (mammals, aves, and herpetofauna) is developed as the classification method to find the suitable candidate species based on the given attributes. The user can select the appropriate species based on the candidate photos displayed on the application. After the user's input is processed, the application will show three species with the most similar features. 95% of the given test cases gave candidates that matched the target. For its ease of use, thus the model is implemented as a mobile application.