Accepted author manuscript, 1.15 MB, PDF document
Final published version
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
}
TY - GEN
T1 - Edge Intelligence-Driven Meteorological Knowledge Graph for Real-Time Decision-Making
AU - Yu, Zheng
AU - Jiang, Jielin
AU - He, Bingkun
AU - Bilal, Muhammad
AU - Liu, Dongqing
PY - 2024/3/26
Y1 - 2024/3/26
N2 - Meteorological decision-making is a crucial element in the meteorological disaster warning and prevention field. With the increasing frequency of meteorological disasters and the rapid development of edge intelligence, there is an urgent need to establish a meteorological early-warning platform that reduces human resource investment, decreases operating costs, and provides targeted information and response suggestions. Therefore, we propose the development of real-time decision-making based on edge intelligence-driven meteorological knowledge graph (EMKG), and aim to achieve meteorological emergency decision-making by combining the knowledge graph with edge intelligence. First, we collect data through edge devices and perform preprocessing and preliminary analysis on these devices to reduce the time and bandwidth requirements for data transmission to the cloud. Based on the above data, meteorological entity recognition and relation extraction were completed using techniques such as BERT, BiLSTM, CRF, and data augmentation. Then we trained a text generation model and deployed it on edge devices to achieve real-time meteorological decision-making. The experimental results show that EMKG effectively integrates edge intelligence and knowledge graph, and further improves the real-time and accuracy of meteorological decision-making.
AB - Meteorological decision-making is a crucial element in the meteorological disaster warning and prevention field. With the increasing frequency of meteorological disasters and the rapid development of edge intelligence, there is an urgent need to establish a meteorological early-warning platform that reduces human resource investment, decreases operating costs, and provides targeted information and response suggestions. Therefore, we propose the development of real-time decision-making based on edge intelligence-driven meteorological knowledge graph (EMKG), and aim to achieve meteorological emergency decision-making by combining the knowledge graph with edge intelligence. First, we collect data through edge devices and perform preprocessing and preliminary analysis on these devices to reduce the time and bandwidth requirements for data transmission to the cloud. Based on the above data, meteorological entity recognition and relation extraction were completed using techniques such as BERT, BiLSTM, CRF, and data augmentation. Then we trained a text generation model and deployed it on edge devices to achieve real-time meteorological decision-making. The experimental results show that EMKG effectively integrates edge intelligence and knowledge graph, and further improves the real-time and accuracy of meteorological decision-making.
KW - knowledge graph
KW - meteorological emergency decision-making
KW - natural language processing
U2 - 10.1109/ICPADS60453.2023.00353
DO - 10.1109/ICPADS60453.2023.00353
M3 - Conference contribution/Paper
AN - SCOPUS:85190249672
T3 - Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS
SP - 2663
EP - 2672
BT - Proceedings - 2023 IEEE 29th International Conference on Parallel and Distributed Systems, ICPADS 2023
PB - IEEE Computer Society Press
T2 - 29th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2023
Y2 - 17 December 2023 through 21 December 2023
ER -