Article Preview
TopIntroduction
Association rules mining is one of the techniques in data mining that extracts interesting but hidden relationships among data objects in a dataset. The initial focus of association rules mining was to explore transaction databases for items frequently purchased together by customers (Mahmood, Shahbaz, & Guergachi, 2014). Modern research has successfully applied the topic in areas such as intrusion detection, telecommunications, disease diagnosis, and education (Mahmood et al., 2014; Abdullah, Herawan, Ahmad, & Deris, 2011). According to Abdullah et al. (2011), two main steps are involved in association rules mining. In the first step, all frequent items are extracted from the transaction dataset. Frequent items are those that appear more than a specified number in the dataset. In the second step, common association rules are generated from the frequent items.
The coronavirus disease 2019 (COVID-19) is a viral infection caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) (Rothan & Byrareddy, 2020; Shereen, Khan, Kazmi, Bashir, & Siddique., 2020). According to Cortegiani, Ingoglia, Ippolito, Giarratano, and Einav (2020), Khan and Atangana (2020), and Shereen et al (2020), COVID-19 was first detected in the city of Wuhan, China in December, 2019. The World Health Organization (WHO) reports that as at 2nd April, 2020, COVID-19 cases had spread to 206 countries, territories and areas; infecting about 896,475 individuals. Out of this figure, a total of 45,525 individuals were reported to have died as a result of the disease (WHO, 2020). So far, there has been no approved vaccine against COVID-19 (Shereen et. al., 2020) whose common symptoms include high body temperature, coughing and problems with breathing.
The objective of this study was to extract association rules from the morbidity and mortality attributes of the novel coronavirus disease 2019. The COVID-19 dataset has five variables that define the morbidity and mortality of the disease. These include confirmed cases, confirmed new cases, confirmed deaths, confirmed new deaths, and number of days since last reported case. The data, as presented by WHO (2020), does not provide information on how the disease variables on morbidity and mortality are associated. To address the gap, this study applies the association rules technique of data mining (Alola & Atsa’am, 2019; Atsa’am, 2020; Bodur & Atsa’am, 2019; Kantardzic, 2009) on the COVID-19 dataset to extract rules that show how the disease morbidity is associated with its mortality. The insights on how the various COVID-19 variables are associated in terms of antecedents and consequents should be useful in the global efforts to tame the pandemic.