You are here:
Guide Data Q&A with Semantic Enrichments
Add semantic statements to help the Data 360 Data Q&A Agent better understand your data's business context, domain-specific terminology, and relationships.
Required Editions
| Available in: All Editions supported by Data 360. See Data 360 edition availability. |
Semantic enrichments are statements you create to provide business and data context to the Data 360 Data Q&A Agent. These statements help the agent interpret your questions more accurately and generate better SQL queries by explaining domain-specific terminology, business rules, data relationships, and calculation methods.
Why Use Semantic Enrichments
While the agent can analyze your data's schema and relationships automatically, it doesn't inherently understand your business context. Semantic enrichments bridge this gap by teaching the agent:
- Business terminology: What "active customer" or "qualified lead" means in your organization
- Calculation methods: How to compute metrics like "customer lifetime value" or "conversion rate"
- Data quality rules: Which fields to use, which to avoid, and why
- Relationships: How objects connect beyond foreign key relationships
- Filtering conventions: Default filters or conditions that should apply to certain queries
How Semantic Enrichments Work
When you ask the agent a question, it considers the applicable semantic enrichments along with the data schema to generate its response. The agent finds the enrichments most relevant to your question based on:
- Object scope (specific object or all objects)
- Field scope (specific field within an object)
- The natural language content of your question
- Lookup summaries that map expected question phrases to the enrichment
Enrichments marked as Critical are always considered. To keep responses accurate, focus each enrichment on a distinct aspect of your data model and avoid contradictory guidance for the same data.
Scope Levels
Semantic enrichments can be scoped at different levels:
- Object
- Applies to a specific Data Lake Object (DLO) or Data Model Object (DMO). Use for object-specific terminology or calculation methods.
- Field
- Applies to a specific field within an object. Use for field-level data quality guidance or interpretation rules.
Examples of Effective Semantic Enrichments
| Use Case | Semantic Statement Example |
|---|---|
| Define business term | "Active customers are those with status = 'Active' and at least one order in the last 90 days." |
| Explain calculation | "Customer lifetime value is calculated as total_revenue divided by months_active, multiplied by average_customer_lifespan_months." |
| Data quality guidance | "When filtering string columns, always use lowercase comparisons because some legacy data has mixed case values." |
| Field interpretation | "The amount_usd__c field represents deal size in US dollars. Use this field for revenue calculations." |
| Default filter | "For opportunity queries, exclude stage = 'Closed Lost' unless explicitly requested." |
Best Practices
- Write statements in clear, plain language that explains context, not just definitions.
- Be specific about when and how rules should apply.
- Mark critical enrichments with the critical flag to ensure the agent prioritizes them.
- Test enrichments by asking the agent questions and reviewing the generated SQL.
- Keep enrichments up to date as your data model and business rules evolve.
Topics
- Create a Semantic Enrichment
Add business context to improve how the Data 360 Data Q&A Agent interprets and responds to questions about your data. - Manage Semantic Enrichments
Edit, deactivate, or delete semantic enrichments as your data model and business rules evolve.
