KnowVID-19: A Knowledge-Based System to Extract Targeted COVID-19 Information from Online Medical Repositories.

Area of Science:

• Biomedical Informatics
• Computational Biology
• Medical Data Science

Background:

• Accessing and synthesizing vast amounts of biomedical literature, particularly concerning rapidly evolving topics like COVID-19, presents significant challenges for researchers.
• Existing search and data mining tools often lack the sophistication to efficiently extract and structure targeted information from diverse online repositories.
• The need for advanced systems to navigate and analyze the growing body of scientific publications is critical for accelerating research and discovery.

Purpose of the Study:

• To introduce KnowVID-19, a novel knowledge-based system designed to enhance the extraction and analysis of information from online medical literature.
• To streamline the process of data extraction and mining for medical researchers and scientists using open-source machine learning tools.
• To improve the efficiency and accuracy of literature searches related to COVID-19 and associated topics through advanced text classification and network visualization.

Main Methods:

• Utilized open-source machine learning tools including GROBID, S2ORC, and BioC for data extraction and mining.
• Implemented a keyword-based text classification process using RAKE, YAKE, and KeyBERT for categorizing publication data.
• Employed the NetworkX Python library to construct and Cytoscape software to visualize networks of relevant terms within publications.

Main Results:

• KnowVID-19 systematically categorizes extracted publication data into distinct topics and subtopics, enhancing query relevance.
• Network visualization of term associations allows for easy tracking of emerging trends in COVID-19 research.
• The system provides an interactive web application with an intuitive interface for seamless keyword searching, filtering, and trend identification.

Conclusions:

• KnowVID-19 significantly improves the efficiency and accuracy of accessing targeted COVID-19 literature for medical researchers.
• The system's topic structuring and network visualization facilitate user-centered exploration and discovery of scientific trends.
• KnowVID-19 offers a valuable tool for navigating complex biomedical information landscapes and accelerating data-driven insights.