Deep Learning Based Fusion and Recommendation of Heterogeneous Multi-Source Intelli-gence for Emergency Events
[Purpose/significance]Emergency events feature complex and diverse information with unpredictable patterns,which is crucial to social and public security.How to quickly and effectively obtain and analyze intelligence in emergencies and carry out safety warnings and crisis intelligence recommendations is an important issue in ensuring social security.[Method/process]This paper con-structs a deep learning based fusion and recommendation system for heterogeneous multi-source intelligence on emergency events.The system utilizes deep learning algorithms to extract features from event data,achieving efficient fusion of heterogeneous multi-source intelligence,and generates and recommends crisis intelligence strategies based on data analytics results.[Result/conclusion]Based on deep learning,a security assurance system framework is constructed,and through empirical analysis,intelligent processing and recommendation of intelligence in complex emergencies are achieved.It effectively improves the efficiency of obtaining and utiliz-ing crisis intelligence in incidents,provides strong support for relevant departments to quickly formulate scientific emergency plans,and has important practical application value.[Innovation/limitation]The study realized the innovative application of deep learning based multi-source heterogeneous information fusion framework in emergency event analysis,which can efficiently acquire intelli-gence and automatically generate recommendations.However,the framework requires a large amount of labeled data for model train-ing.The workload of manual annotation is extensive,and the consistency and correctness of sample labeling need further improvement.
deep learningheterogeneous multi-source intelligencedata fusioncrisis intelligence recommendationintelligence gen-eration
NETL
NSTL
万方数据
