摘要
在人工智能与大数据技术深度融合的时代背景下,公安机关在利用开源情报(OSINT)进行异常行为分析时,面临数据源扩张和数据规模激增带来的挑战,传统检测方法已难以有效应对复杂检测任务。对此,应构建多模态开源情报异常检测框架,通过边缘计算与云计算协同机制优化数据传输效率,利用自适应采样策略与主动学习机制破解数据标注难题,引入深度学习模型实现图像去噪、音频降噪及文本语义增强的效果,以提升公安部门对异常行为的预警响应与应急处置能力,为维护社会稳定提供关键的技术支撑。
In the era of deep integration between artificial intelligence and big data technologies,public security organs face challenges from the expansion of data sources and the surge in data scale when using open source intelligence(OSINT)for abnormal behavior analysis.Traditional detection methods are no longer effective in dealing with complex detection tasks.In this regard,we should build a multi-modal open source intelligence anomaly detection framework,optimize the data transmission efficiency through the edge computing and cloud computing collaborative mechanism,use adaptive sampling strategies and active learning mechanisms to solve the problem of data annotation,and introduce the deep learning model to achieve the effect of image denoising,audio denoising and text semantic enhancement,so as to improve the early warning response and emergency response capabilities of the public security department to abnormal behaviors,and provide key technical support for maintaining social stability.
作者
林逸飞
张璐
Lin Yifei;Zhang Lu(Shandong Police College,Jinan,250200,Shandong)
出处
《湖南警察学院学报》
2025年第4期14-24,共11页
Journal of Hunan Police Academy
基金
2022年山东公安科技创新项目“人脸深度伪造视频检测技术研究”(GAKJCX2022-8)。
关键词
机器学习
开源情报
异常行为检测
多模态数据
machine learning
open source intelligence
abnormal behavior detection
multi-modal data
