A collaborative effort by Southern University of Technology (SUSTech) and the University of Macau (UMacau) has seen the development of mining frequent patterns over the 2019-nCOV knowledge graph.
SUSTech Assistant Professor Tang Bo and UMacau Professor Leong Hou U have created a system that allows users to discover frequent patterns in the 2019-nCOV knowledge graph. The outputs could assist decision-makers in making better choices when it comes to dealing with the novel coronavirus (COVID-19).
The rapidly evolving novel coronavirus (COVID-19) situation has seen a vast increase in the volume of data available to decision-makers, with the types of data increasing at an exponential rate. Working out how to efficiently use this information is an enormous challenge, so mining these for frequent patterns is vital to improving the effectiveness of clinical treatment.
The frequent pattern mining system establishes a knowledge graph index structure after it processes each new batch of novel coronavirus data. The index structure supports efficient frequency pattern mining & allows for user input of their requirements. Preliminary research results indicate calculation efficiency 10 to 20 times faster than the state-of-the-art algorithms. This new system also allows for significantly more complex searches combined with real-time data processing.
The National Natural Science Foundation of China and the Shenzhen Basic Research Free Exploration Project supported the core technology source of the computing layer.
The system can analyze different novel coronavirus knowledge graphs. It has looked at mutation patterns of COVID-19, providing information about the number of strains in various cities around the world. This sort of analysis helps experts understand the characteristics of the novel coronavirus (COVID-19) in cities across the globe and better determine optimal clinical treatment options.
Assistant Professor Tang Bo said that “the existing technology struggled to meet the needs of experts to analyze the spread of the novel coronavirus accurately. This system allows them to query and analyze the 2019-nCOV knowledge graphs so they can make better decisions.”
The data comes from the Chinese Open Knowledge Graph novel coronavirus group. The novel coronavirus Knowledge Graph is based on the unified naming specification and semantic format and adopts CC by SA similar signature open license agreement.
Doctoral candidate Zeng Jian, class of 2022 undergraduate Tang Qiandong and class of 2020 undergraduate Yang Chuan, were the main contributors to the system. It will continue to be refined through the advice of experts.
