• Wilfrid Laurier University

• CP 421

Exploring Data Mining Techniques in COVID-19 Research

Alexandros Ioannou#1, Cameron Anderson$2, Lubna Al Rifaie*3, Luke Aikman^4

1[email protected], 2[email protected], 3[email protected], 4[email protected]

Abstract—The COVID-19 pandemic posed unprecedented and

unpredictable challenges across the world, requiring the

deployment of complex data mining techniques to manage and

mitigate the impact of the virus. This paper reviews recent

research works that apply various data mining methods, such as

Natural Language Processing (NLP), supervised learning

techniques, clustering algorithms, frequent itemsets, and

association rules, to address the big data challenges within the

healthcare sector during the COVID-19 crisis. By analyzing these

methods’ applications in tracking the virus spread, predicting

outcomes, and enhancing healthcare responses, we aim to identify

existing data mining problems and propose viable solutions. The

insights from this study show the potential of data mining in

revolutionizing healthcare monitoring and information

technology, particularly in managing pandemic situations.

Keywords—Data mining(DM), COVID-19, Healthcare

monitoring, Clustering algorithms, Frequent itemsets,

Association rules, Big Data, Predictive analytics.

I. INTRODUCTION

The emergence of the COVID-19 pandemic has not only

brought up a global health crisis but also led to an increased

dependence on data science and informatics in public health

responses. The vast amounts of data produced by health

monitoring systems, contact tracking, and case reporting offer

both opportunities and challenges. Understanding the

complexity of COVID-19 data has become more dependent on

data mining techniques, which have historically been essential

in extracting relevant information from huge datasets. To

support well-informed healthcare decisions and public health

policies this study explores the various applications of these

techniques – ranging from NLP and supervised learning to

association rules – in filtering through data linked to

pandemics.

One of the methods in the reviewed papers was the use of

Natural Language Processing. This method was a key tool for

tracking misinformation, analyzing sentiment on social media,

and compiling clinical data to comprehend the dynamics of the

pandemic. Similarly, significant effectiveness was shown by

supervised learning algorithms in forecasting infection rates,

and patient outcomes, and identifying high-risk groups, which

allowed for more focused interventions. Also, clustering

algorithms have made it possible to stratify patient groups and

implement individualized treatment plans by providing

insights into patient symptomatology and disease

development.

However, the deployment of data mining in healthcare,

particularly in a crisis of this magnitude, is full of challenges.

Data quality and availability remain significant obstacles, with

the heterogeneity of data sources and formats complicating