Intelligent Classification and Tagging enables you to identify sensitive data and manage business tags through natural language descriptions. Buddy automatically scans field characteristics, recommends governance tags, and waits for your confirmation at key points before applying them in batches.
Overview
Data classification and tagging is a core step in data governance that connects asset discovery and security policies. By classifying and tagging tables and fields, you can provide a foundation for subsequent security policy configuration and quality rule recommendations. This feature covers the following scenarios:
Automatic sensitive field identification: scans table field names to automatically identify PII fields such as mobile numbers, ID card numbers, email addresses, and bank card numbers, and matches them with built-in sensitive tags. If a user has previewed the table data in Catalog, AI also performs identification based on the actual data content (for example, determining whether a column's values conform to the mobile number format) to improve identification accuracy. If the data has not been previewed, inference is performed based only on field names.
Custom tag support: in addition to system built-in tags, users can define custom governance tags. The Agent can perform identification and recommendation based on the custom tag system to meet business-specific sensitive field identification requirements.
Business tag recommendation: recommends appropriate business tags based on table names, field semantics, and the business context of the associated Schema (such as domain:finance and tier:gold).
Batch Tagging: Applies recommended governance tags to multiple tables and fields in batches at once, with no limit on the number of tags.
Compliance Identifier Association: For identified sensitive fields, AI recommends associating compliance framework identifiers such as PII, GDPR, and PIPL in the conversation, and performs the association only after confirming each item with the user.
Prerequisites
Buddy has been enabled for the workspace.
The current account has Buddy development mode permissions.
You have view permissions on the Catalog / Schema to be tagged.
To write tags, you need the tag management permission on the corresponding Catalog.
Information You Must Provide
Before performing classification and tagging, Buddy will ask you to provide the following minimum information set, while the rest is automatically recommended by AI:
|
Tagging scope | Target Catalog, Schema, or specific table name |
Tagging purpose | Sensitive identification / Business tag recommendation / Compliance tag association (can be implicitly expressed through natural language) |
If your input is incomplete, Buddy will ask for the missing information through an "Ask User card" before continuing.
Content That AI Can Intelligently Generate
|
Sensitive Field Identification Result | When data is not previewed, inference is based on field name patterns (such as phone, id_card, email); when data has been previewed, matching is performed based on actual data content characteristics. |
Sensitive Tag Recommendation | Match the system built-in tag library (PII, finance, healthcare, and other categories) |
Custom Tag Matching | AI identifies matching fields based on user-created custom tag definitions. |
Business Tag Recommendation | Analyze table names, Schema names, field semantics, and business context, and recommend business tags such as domain and tier. |
Compliance Framework Association | Recommend associating compliance identifiers such as PII / GDPR / PIPL based on sensitive field types, which requires user confirmation in the conversation. |
Operation Steps
Step 1: Expressing Classification and Tagging Requirements
Describe your requirements in natural language in the Buddy input box, and Buddy will automatically identify the tagging type and target scope.
Example Input:
Scan all fields of ods_catalog.user_info and help me identify which ones are sensitive fields.
Tag all tables under the dw_order Schema with the domain:ecommerce business tag.
Which fields in this table require attention for PIPL compliance?
Tag all these sensitive fields with the system built-in tags.
Based on our custom "Business Sensitivity Classification" tag system, scan the sensitive fields under user_domain.
Step 2: Completing Missing Information
If the input is incomplete, Buddy will ask you for the missing information through an Ask User card:
If no Catalog / Schema / table is specified, the available scopes are listed for you to choose from.
If custom tags are used but no tag system is specified, the created tag systems are listed for you to choose from.
After you answer, Buddy will continue the execution.
Step 3: Reviewing AI Recognition and Recommendation Results
Buddy displays the scan results in a table:
Sensitive Field Identification:
Field name, table, and field type
Identified sensitive types (phone number, ID card, email, bank card, and so on)
Recommended sensitive tags
Match confidence
Identification basis (field name inference, data content recognition)
Business Tag Recommendations:
Table name and field name
Recommended tags and reasons
Existing tags (to avoid duplicate tagging)
Note:
You can modify, delete, or add tags in the recommendations before confirming.
Step 4: Confirming and Tagging in Batches
After you confirm the tag recommendations, Buddy calls the DataBuddy API to write tags in batches and displays:
Number of successfully tagged tables and fields
Failed items and reasons (such as insufficient permissions or missing tags)
There is no limit on the number of tags that can be applied in a batch.
Step 5: Confirming Compliance Label Association
For identified sensitive fields, Buddy recommends associating compliance framework identifiers such as PII / GDPR / PIPL one by one in the conversation:
AI displays the recommended compliance identifiers for each sensitive field.
You can confirm or reject each item in the conversation.
After confirmation, Buddy performs the association operation and displays the results.
Step 6: Integrating Subsequent Governance (Optional)
After completing sensitive data identification and tagging, Buddy proactively prompts whether to proceed with subsequent governance operations:
|
Configuration for the security policies. | |
Compliance Coverage Check | |
Key Milestone Confirmation
|
Scan scope | Before Scanning | Catalogs / Schemas / Tables to scan |
Tag system | When Using Custom Tags | Confirm which tag definition to use. |
Recognition result | Before Tagging | Review sensitive identification and tag recommendations, which can be modified. |
Batch tagging | Before Write | Confirm the impact scope and tag list. |
Compliance label | Before Association | Confirm compliance framework identifier associations item by item in the conversation. |
FAQs
Q: What Is the Basis for Identifying Sensitive Fields?
A: The identification basis depends on whether you have previewed the table data in Catalog. If the data has not been previewed, Buddy infers based only on field name patterns (for example, phone → mobile number, id_card → ID card number). If the data has been previewed, AI also performs identification based on actual data content characteristics (for example, determining whether a column's values conform to the mobile number format), resulting in higher accuracy.
Q: Does Batch Tagging Overwrite Existing Tags?
A: No. By default, Buddy uses the "append" mode, which only adds recommended tags without deleting existing ones. If a recommended tag conflicts with an existing tag, a warning will be marked in the results.
Q: How Do I Create a Custom Tag?
A: Create custom tag categories and tag values in Governance Tag Overview. After creation, Buddy can use these custom tags for matching during scanning. Q: Is Compliance Label Association Automatic or Manual?
A: Confirmation is required. Buddy recommends associating compliance framework identifiers such as PII / GDPR / PIPL in the conversation, and performs the association only after confirming each item with you. It does not associate automatically.
Q: What Is the Relationship Between Sensitive Fields Identified by Buddy and Built-in Identification Rules of the Data Security Module?
A: Buddy's sensitive identification results are written to the governance tag system, complementing the built-in identification rules of the data security module. Sensitive fields identified by both are displayed in a unified manner in Catalog.
References