\nT/GXDSL 026-2025 Artificial Intelligence Data Annotation Service Specification

Scope

Main technical content: This standard specifies the basic requirements, process norms, quality control, security and privacy protection, and personnel management for AI data annotation services. This standard applies to all enterprises, institutions, and individuals that provide AI data annotation services, as well as users who utilize data annotation services. Specifically, it includes but is not limited to the following fields: - Computer vision: image classification, object detection, image segmentation, image annotation, video analysis, etc. - Natural language processing: text classification, sentiment analysis, machine translation, question-answering systems, text summarization, etc. - Speech recognition: speech-to-text conversion, speech synthesis, speech emotion analysis, etc. - Data mining: data cleaning, data classification, data clustering, association rule mining, etc. - Other fields: medical imaging analysis, autonomous driving, intelligent customer service, financial risk control, etc. This standard applies to the following scenarios: - Internal data annotation within enterprises: Enterprises conduct data annotation for their own business needs. - Outsourced data annotation services: Enterprises outsource data annotation tasks to professional data annotation service providers. - Crowdsourced data annotation: Distribute data annotation tasks to a large number of users via crowdsourcing platforms. - Open-source data annotation: Open data annotation platforms for users to freely use and contribute data. This standard does not apply to the following content: - Data annotation involving state secrets. - Data annotation involving personal privacy, unless authorized by relevant authorities. - Other data annotation activities prohibited by laws and regulations.

Development Information

Related Standards