面向公共安全的多源异构数据结构化提取关键技术研究

Research on Key Technology of Multi-source Heterogeneous Data Structured Extraction for Public Security

提出了多源异构数据智慧驱动的平台架构,运用深度神经网络、自然语言处理等多项人工智能技术,将声音、图像、视频和文字等多种非结构化数据中蕴藏的丰富信息提取为结构化数据,辅以其他数据,为城市应急、防控、维稳等领域提供技术保障。

This paper proposes a multi-source heterogeneous data intelligence-driven platform architecture.Using deep neural networks,natural language processing and other artificial intelligence technologies,the rich information contained in a variety of unstructured data such as sound,image,video and text is extracted into structured data,supplemented by other data,to provide technical support for the fields such as emergency response for the city,prevention and control and stability maintenance.