数据安全分类分级数据集合

采用基于BERT的文本特征提取模型及CNN图像特征提取模型，自动识别数据中的关键信息，提取特征标签，为分类分级提供数据支撑。分级权重计算算法：基于层次分析法（AHP）构建分级指标体系，将“数据影响范围”、“数据价值密度”、“泄露后果严重程度”作为一级指标，通过该数据集合内置的行业权重数据库，自动为各指标分配权重，最终通过加权求和得出数据级别。A BERT-based text feature extraction model and a CNN-based image feature extraction model are adopted to automatically identify key information in data, extract feature tags, and provide data support for classification and grading. The grading weight calculation algorithm constructs a grading index system based on the Analytic Hierarchy Process (AHP), takes "data impact scope", "data value density" and "severity of leakage consequences" as primary indicators, automatically assigns weights to each indicator through the industry weight database built into this dataset, and finally obtains the data level through weighted summation.