```json { "classifications": [ { "column": "batch_id", "role": "identifier", "confidence": 1.0, "reason": "Primary key with unique values per row (304 distinct in 304 rows), used to identify batches" }, { "column": "dataset_id", "role": "identifier", "confidence": 1.0, "reason": "Foreign key referencing datasets, used for JOINs (96 distinct values, repeating across batches)" }, { "column": "tenant_id", "role": "dimension", "confidence": 1.0, "reason": "Categorical field for grouping/filtering by tenant (only 1 value 'default' currently, but designed for multi-tenancy)" }, { "column": "started_at", "role": "dimension", "confidence": 1.0, "reason": "Timestamp for filtering and grouping by time periods (304 distinct timestamps)" }, { "column": "completed_at", "role": "dimension", "confidence": 1.0, "reason": "Timestamp for filtering and grouping by completion time (285 distinct values, some nulls for incomplete batches)" }, { "column": "status", "role": "dimension", "confidence": 1.0, "reason": "Categorical field with 3 values (completed, ready, running) used for filtering and grouping" }, { "column": "error_text", "role": "attribute", "confidence": 1.0, "reason": "Free-form descriptive text for displaying error messages (100% null currently, but designed for display)" }, { "column": "total_rows", "role": "measure", "confidence": 1.0, "reason": "Numeric quantity to aggregate (sum/avg) representing row counts processed" }, { "column": "total_bytes", "role": "measure", "confidence": 1.0, "reason": "Numeric quantity to aggregate (sum/avg) representing bytes processed" }, { "column": "elapsed_ms", "role": "measure", "confidence": 1.0, "reason": "Numeric quantity to aggregate (sum/avg) representing processing time in milliseconds" }, { "column": "engine_versions_json", "role": "dimension", "confidence": 0.9, "reason": "Low cardinality (7 distinct values) version information used for grouping/filtering by engine configuration" }, { "column": "nettle_version", "role": "dimension", "confidence": 1.0, "reason": "Low cardinality categorical field (2 values) for grouping/filtering by version" }, { "column": "job_id", "role": "identifier", "confidence": 1.0, "reason": "Foreign key or alternate identifier (295 distinct values) used for JOINs to job records" }, { "column": "content_hash", "role": "identifier", "confidence": 0.95, "reason": "Hash code used to identify unique content (85 distinct values, 55% null), serves as deduplication key" } ] } ```