Crosswalks for Interoperability
Definition
A crosswalk is an explicit mapping between two classification systems, enabling data coded in one system to be translated into another.
Why This Matters
No single classification system serves all purposes. The SOC exists alongside O*NET (which adds skill and ability data), NAICS (industry classification), CIP (educational programs), and international systems like ISCO. Each system optimizes for different use cases, but analysis often requires connecting data across them.
Without maintained crosswalks, every integration project reinvents mappings, introducing inconsistencies and errors. With them, interoperability becomes a lookup rather than a research project.
How It Works
Crosswalks can be:
- One-to-one: Category A maps to Category B
- One-to-many: Category A maps to Categories B, C, D (with proportions or conditions)
- Many-to-one: Categories A, B, C all map to Category D
- Many-to-many: Complex mappings requiring conditional logic
The SOC publishes crosswalks to Census occupation codes, O*NET-SOC codes, previous SOC versions (2010 → 2018), and international systems. These aren’t just reference documents, they’re infrastructure that enables cross-system analysis.
Implications for Knowledge Engineering
When designing classification systems, plan for crosswalks from the start:
- Document the semantic intent behind each category (not just the label)
- Track which categories are “clean” mappings vs. require splitting or combining
- Maintain crosswalk provenance: who created it, when, and under what assumptions
- Version crosswalks alongside the classification systems they connect
A classification system without crosswalks is an island. The value often lives in the connections.
Related: 04-atom—time-series-continuity-constraint, 02-molecule—hierarchical-aggregation-levels, 06-atom—direct-match-concept