Interesting Patterns From COVID‐19 Dataset Using Graph‐Based Statistical Analysis for Preventive Measures
C B Abhilash, Kavi Mahesh · 2021
The coronavirus disease (COVID-19) caused by the novel severe acute respiratory syndrome–coronavirus 2 (SARS-CoV-2) was declared a global pandemic on March 11, 2020, by WHO. Different countries adopted many interventions at different levels of the outbreak. The common one is social distancing and lockdown, which were opted to flatten the mortality curve [13]. The important aspect of data mining is knowledge discovery for interesting patterns. We propose an innovative approach for deriving interesting patterns using machine learning and graph database models to incorporate the preventive measures in an earlier state. We propose a graph-based statistical analysis (GSA) model to study the pandemic outbreak's impact and propose prevention and control strategies. We analyze the model using available data from the ministry of health and family welfare service, GOI. The study will explore the visual analytics for deriving the interesting patterns from the knowledge graph considering the attributes of infected COVID-19 patients, which will benefit the preventive measures. The pandemic's spread to the respective geographical population is predicted using the time series and statistical analysis with the statistical model. This method often requires sufficient data, but with the available data about the initial and mid-stages of an epidemic, the spread rate is predicted for preventive measures that the administrative levels should take. Our study uses Neo4j, a graph database tool for building the knowledge graph represented by RDF (Resource Descriptive Framework) for optimized query operations resulting in graph visualization. On the other hand, we also discuss the machine learning model to predict the trend and short-term transmission predictions through the COVID-19 outbreak. The results have potentially interesting insights on the effect of the pandemic.