Distributed Classifier for SDGs Topics in Online News using RabbitMQ Message Broker

Abstract

Internet data has grown very fast and becomes very large. Thus continuous improvement will always be required to face this challenge. The Sustainable Development Goals (SDGs) are defined by the United Nations (UN) to encourage improvements in the field of life in each country. We proposed a combination of Distributed System (RabbitMQ) and Machine Learning (Naïve Bayes Classifier) as one of the support to measure the achievement level of Sustainable Development Goals (SDGs) in Indonesia. The methods will categorize the Detik.com news into two classes; the relevant to SDGs of Indonesia and the irrelevant to SDGs of Indonesia. Our work shows that the use of the load-balance feature in RabbitMQ could shorten the processing time of the Naïve Bayes Classifier. RabbitMQ as a load-balancer can divide the workload equally, thus reducing the latency time of the Naïve Bayes Classifier classification process by 30.3 percent.