Skip to content

회원가입 시 AI 기반 악의성 검열 로직 적용 #126

@ahyeonkong

Description

@ahyeonkong

📄 설명

AiModerationService는 채팅, 회원가입 등 여러 API에서 공통으로 사용할 수 있습니다.

악의성 판단 기준

  1. is_harmful 필드로 true/false 값 판별
  2. confidence 필드로 필터링 세분화("매우 위험”, “위험”, “보통” 등)

가장 기본적인 첫 번째 방법으로 진행합니다.

✅ 작업할 내용

  • AI 응답 객체 생성(AiModerationResponse)
  • AI 요청 서비스 생성(AiModerationService)
  • AI URL 추가(로컬/배포)
  • 회원가입 로직에 AI 검열 추가

🙋🏻 참고 자료

Metadata

Metadata

Assignees

Labels

No fields configured for Feature.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions