A Data Mining and Ontology-Based Approach for Predicting the Research Ideas in the Wildlife Sector of Sri Lanka

Abstract

Sri Lanka, is the global biodiversity hotspot, which contains a wide variety of fauna & flora. Also, wildlife is the key player in the tourism industry of Sri Lanka. But still, Sri Lankan wildlife faced many conflicts to protect their biodiversity because of a lack of technical and research support. There were significant research gaps and data resources but still, researchers retreat to select this domain as their study. This study demonstrates a novel approach to data mining to find hidden keywords and automated labeling for past research work in this area. To model topics and recognize the major keywords for future researchers, we used Latent Dirichlet Allocation (LDA) algorithms to build an ontology model to visualize the relationships between each keyword and similar keywords cluster into creative topics. These techniques are also useful for potential research ideas, for identifying hidden research gaps and to classify domain-related publications. The experiment results demonstrate the validity and efficiency as well as it will make a platform for all researchers to interact easily with the wildlife domain.