A text classification on the downstreaming potential of biomedicine publications in Indonesia

Abstract

This study has the purpose to investigate the potential to downstreaming of biomedicine researches in Indonesia based on scientific publications. It is therefore necessary to extract unstructured information in natural language-based scientific publications. This paper reports result from an investigation on a classification model of the downstreaming potential of biomedical research publications in Indonesia based on text-mining. The predictive computational model was built by testing three classifier algorithms namely KNN, Naive Bayes and SVM, where the results show that the Naive Bayes-based model performs best.