A Data Visualization Framework during Pandemic using the Density-Based Spatial Clustering with Noise (DBSCAN) Machine Learning Model

Abstract

Big data technologies have become an important part in our life, especially during the pandemic. These technologies can be used to collect, analyse, process, and interpret the collected data in order to produce a useful information or knowledge. In fact, we are depending on the information extracted from the large amount of data collected daily from mobile applications. One of the examples of the application that has been used in Malaysia is MySejahtera which provides useful information on the spread of the pandemic. The data can be clustered using machine learning models such as clustering algorithm. Therefore, in this project, we propose a framework that will be useful to monitor the information about COVID-19 and visualizing the information with a machine learning model. The data visualization can help with data interpretation and improving how we can manage the spread of the virus. This project was also implemented using a modified waterfall which allows the developer to return to the previous phase in order to make some modifications before the final product can be used by users. This project used a Python approach to develop a dashboard. A Density-Based Spatial Clustering with Noise algorithm was chosen for the data classification of the countries based on its number of cases and number of deaths.