Evolutionary Algorithm Based Summarization for Analyzing COVID-19 Medical Reports

Mallick, C.; Das, S.; Das, A. K.

The COVID-19 pandemic emerged in the month of December from the very popular city Wuhan of China and spread rapidly all over the country. The present situation depicts the progress in various less-explored areas to strengthen the global knowledge base for preventing and addressing the COVID-19 pandemic. Novel coronavirus disease became a major health concern due to its pathogenicity and world widespread distribution. Many good scientific research papers related to COVID-19 have recently published using the dataset CORD-19. Analysis of these massive data and identifying meaningful information is a crucial and very challenging task in health care and bio-medical science research. In this chapter, we try to extract important information about COVID-19 from the available text documents, such as research papers, articles, journals, reports, and other publications. Initially, in the chapter, these types of text data are preprocessed by means of sentence tokenization, then removal of stopwords, and finally stemming operations. Consequently, a bag-of-words is generated to represent each report in a structured form and a bi-objective strength Pareto evolutionary algorithm based text summarization technique is concerned to produce the summary of the report. The method has two objective functions: one based on the length of the summary, a predetermined value, and the other based on the clustering coefficient concept of graph theory. After convergence of the evolutionary algorithm, the fittest chromosome in the final population provides a summary of the report. The method is applied on COVID bio-medical reports publicly available in CORD-19 dataset for generating a summary and its performance is finally evaluated by comparing it with a few of the related state-of-the-art methods. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.