← 返回主平台

审计与检查场景

快速处理审计数据，保护隐私，输出审计结论

场景描述

审计团队需要处理交易明细、客户清单、操作日志等大量数据。通常流程：

  1. 获取原始交易数据（包含敏感字段）
  2. 保护个人信息（脱敏）
  3. 修正数据异常和缺失
  4. 按审计维度分组统计
  5. 生成审计检查口径

处理方案

步骤 1：脱敏敏感字段

[
  {
    "type": "mask",
    "preset": "name",
    "field": "customer_name",
    "keepFirst": 1
  },
  {
    "type": "mask",
    "preset": "idcard",
    "field": "id_number"
  },
  {
    "type": "mask",
    "preset": "phone",
    "field": "phone_number"
  }
]

步骤 2：清洗和规范化

[
  {
    "type": "normalize",
    "field": "transaction_date",
    "mode": "date"
  },
  {
    "type": "normalize",
    "field": "amount",
    "mode": "number"
  },
  {
    "type": "fillMissing",
    "field": "transaction_type",
    "strategy": "default",
    "defaultValue": "unknown"
  }
]

步骤 3：计算审计指标

[
  {
    "type": "compute",
    "targetField": "is_large_amount",
    "formula": "IF(amount > 100000, 'yes', 'no')"
  },
  {
    "type": "compute",
    "targetField": "transaction_month",
    "formula": "DATE_FORMAT(transaction_date, 'YYYY-MM')"
  }
]

步骤 4：聚合检查结果

{
  "type": "aggregate",
  "groupBy": ["transaction_month", "is_large_amount"],
  "targets": [
    { "field": "customer_id", "agg": "count", "outputField": "transaction_count" },
    { "field": "amount", "agg": "sum", "outputField": "total_amount" },
    { "field": "amount", "agg": "avg", "outputField": "avg_amount" },
    { "field": "amount", "agg": "max", "outputField": "max_amount" }
  ]
}

关键指标

指标 用途 生成方式
交易笔数 核实交易总量 COUNT(transaction_id)
交易总额 核实金额合计 SUM(amount)
平均交易额 判断异常 AVG(amount)
超大交易笔数 风险控制 COUNT(IF amount > threshold)

最佳实践

  • ✅ 先脱敏，再分享或导出数据
  • ✅ 规范化日期和数字格式，确保准确
  • ✅ 使用多维度分组（如日期 + 金额等级）
  • ⚠️ 避免在聚合前删除原始数据，留作对账
💡 相关指南： 脱敏规则 | 清洗规则 | 聚合规则