Data Standard Management Practices in Ctrip Vacation Data Governance

Data governance aims to increase data value through a comprehensive management system that includes organization, policies, processes, and tools; the article begins by explaining why data standard management is essential as the foundation of effective governance.

The three core elements of data standard management are defined as Scope (clearly identifying what needs governance), Tools (systematic support for standard activities), and Policies (guidance for people and processes).

1. Data Integration Practice

The Ctrip Vacation platform handles structured, semi‑structured, and unstructured sources. Production change standard management addresses two main questions: which changes need notification and who should be notified. Structural changes (DDL) are automatically detected and reported with details such as database, table, owner, change type, and object. Content‑level changes require manual impact assessment and offline communication. The system also supports online manual reporting for enumerated value changes, field deprecation, logic changes, table deprecation, and migration.

Impact analysis leverages a data lineage component that records relationships between tables, fields, and owners, enabling automated identification of affected parties and downstream applications. Future work will extend this analysis to application‑level lineage.

2. Metadata Management Practice

Metadata is treated as a product; a data map aggregates four types of business metadata: models, metrics, dashboards, and datasets, each with standardized naming, ownership, and access controls.

Table metadata standards cover layering (ODS, EDW, CDM, ADM, MID, DIM), primary and secondary business domains, partitioning, importance level (P0‑P3), and sensitivity (L1‑L4). A tooling system generates DDL templates that enforce naming conventions and comments.

Metric metadata standards define business terms, data domains, processes, time cycles, modifiers, atomic and derived metrics, and dimensions, ensuring consistent definition and automated generation of metric lineage.

3. Standardization of Portal Dashboards and Self‑Service Queries

Portal dashboards enforce the use of standardized metric IDs, displaying clear definitions and linking to detailed metadata pages to avoid ambiguous naming. Self‑service query templates provide pre‑defined, well‑commented SQL snippets that business users can customize with parameters for one‑click data extraction.

Conclusion

Effective data governance requires coordinated organizational effort, clear standards, automated tooling, and continuous quality feedback loops. Ctrip Vacation’s experience demonstrates how systematic standard management, robust metadata practices, and standardized consumption interfaces can improve data reliability and business value.