Data Models and Schema Changes

字数

566 字

阅读时间

3 分钟

New Data Models

SuggestionFeedback

• Purpose: 记录用户对 AI 微建议的即时反馈，支撑模型调优与个性化排序。
• Integration: 与现有 suggestions、tasks 集合关联，形成“建议 → 反馈 → 任务执行”闭环。
• Key Attributes:
  • suggestion_id: string – 对应的微建议标识
  • user_id: string – 用户标识
  • feedback_type: string – 反馈类型 (helpful/neutral/not_helpful)
  • feedback_score: number – 1-5 满意度评分
  • comment: string – 可选文字说明
  • action_taken: string – 是否转化为任务、收藏或忽略
  • createdAt: date, updatedAt: date
• Relationships: With Existing: suggestions, tasks. With New: TaskExecutionLog.

TaskExecutionLog

• Purpose: 精细记录任务执行过程、打卡记录与复盘数据，满足 PRD 中对任务闭环的统计与周报需求。
• Integration: 与 tasks、checkins、reports 集成，为周报和徽章系统提供详尽数据。
• Key Attributes:
  • task_id: string – 任务唯一标识
  • user_id: string – 用户标识
  • event_type: string – 执行事件 (start/complete/review)
  • event_payload: object – 事件详情
  • linked_checkin_id: string – 对应情绪回执
  • effectiveness_score: number – 执行后自评效果 0-1
  • createdAt: date, updatedAt: date
• Relationships: With Existing: tasks, checkins, reports. With New: ModerationAuditTrail.

ModerationAuditTrail

• Purpose: 为树洞风控与 SOS 保障建立完整审核日志，满足 PRD 对“风险识别 → 审核 → 干预”链路的审计要求。
• Integration: 与 posts, comments, sos_events, notifications 交互。
• Key Attributes:
  • target_type: string – 被审核对象类型
  • target_id: string – 对象 ID
  • risk_snapshot: object – 风险指标
  • action_taken: string – 采取的措施
  • reviewer_id: string – 审核人
  • follow_up_task_id: string – 后续干预任务
  • createdAt: date, updatedAt: date
• Relationships: With Existing: posts, comments, sos_events, notifications. With New: TaskExecutionLog.

Schema Integration Strategy

• New Tables: suggestion_feedback, task_execution_logs, moderation_audit_trail.
• Modified Tables:
  • suggestions: add feedback_count, last_feedback_at.
  • tasks: add latest_execution_score.
  • posts: add last_reviewed_at, audit_status.
  • checkins: add mood_intensity, energy_level.
• New Indexes: For new collections and updated fields to support queries.
• Migration Strategy: Use cloud function batch processing to add default fields. Create mock_datasets collection to centralize future mock data.
• Backward Compatibility: Ensure old data defaults gracefully and APIs tolerate missing new fields.

贡献者

Cai Hongyu

文件历史

最后编辑于 4 个月前查看完整历史

fbb52-2025-10-31 10:24:10