Discover the Trends and Hotspots of Biosafety and Biosecurity Research via Machine Learning

Renchu Guan^1,2, Haoyu Pang¹, Yanchun Liang^1,2, Zhongjun Shao³ , Xin Gao^4,5 , Dong Xu⁶ , Xiaoyue Feng^1,*

1 Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, College ofComputer Science and Technology, Jilin University, Changchun, 130012, Jilin, China,
2 Zhuhai Laboratory ofKey Laboratory of Symbolic Computation and Knowledge Engineering of the Ministry of Education, ZhuhaiCollege of Jilin University, Zhuhai, 519041, Guangdong, China,
3 Department of Epidemiology, Ministry ofEducation Key Laboratory of Hazard Assessment and Control in Special Operational Environment, Schoolof Public Health, Air Force Medical University, Xi’an, 710032, Shaanxi, China,
4 Computational BioscienceResearch Center (CBRC), Computer, Electrical and Mathematical Sciences and Engineering (CEMSE)Division, King Abdullah University of Science and Technology (KAUST), Thuwal, 23955-6900, Saudi Arabia,
5 BioMap, 100000, Beijing, China,
6 Department of Electric Engineering and Computer Science, and Christopher S. Bond Life SciencesCenter, University of Missouri, Columbia, 65201, Missouri, USA

* Corresponding author, E-mail: fengxy@jlu.edu.cn

Abstract

Up to September 3,2021, 219,927,106 cases of Corona Virus Disease 2019 (COVID-19) have been reported worldwide (4,552,873 deaths), and there is no end trend. The problems reflected in this epidemic have made biosafety and biosecurity (hereafter collectively referred to as "biosafety") once again a hot topic in the international community. And biosafety research covers a wide range and diversity, and it is very important to quickly retrieve biosafety hot-spots and trends from massive data. We design and construct a novel model LDAPR, which is combined with the Latent Dirichlet allocation model, affinity propagation clustering, and PageRank algorithm to extract topics from the abstracts of biosafety-related papers from 2011 to 2020. And then we conduct hotspots and trends analysis on the results of LDAPR and carry out further research on the connections among these topics, include annual hot topic extraction, ten-year keyword evolution trend analysis, topic map construction, hot spot area discovery, fine-grained correlation analysis between topics, interdisciplinary research trends, etc. Finally, all experimental results and analyses are visualized in various ways, like the word clouds, sequence diagram, chord diagram, heat map, radar diagram, hot spot world map, and bar plot. Relevant research results prove that these analyses can provide inspiration for further research. Detailed experimental results can be obtained from www.keaml.cn/Biosafety