Database Key Relationship Determination via Sampling

The patent segments the entire database table into multiple sampled data sets by dividing rows into batches. Instead of querying the full table, SQL queries are executed on these smaller segments. The sampled data sets are processed in parallel or sequentially to determine composite key relationships, significantly reducing query execution time while maintaining acceptable accuracy through statistical sampling methods.

The patent applies partial action by querying only a sampled portion of the database table rather than the complete data set. By selecting representative samples (e.g., 1000-10000 rows) from the full table, the system obtains sufficient information to determine key relationships without the overhead of processing all rows, thus reducing time consumption while preserving measurement precision.

The patent divides large database tables into manageable sampled data sets that can be processed efficiently. By segmenting the data and applying parallel processing techniques, the system maintains productivity even as dataset size increases. The segmented approach allows multiple processors to work simultaneously on different portions of the data, preventing productivity degradation with larger tables.

The patent changes the parameter of data volume from complete table size to sampled subset size. By adjusting sampling parameters (such as sample size, batch size, and confidence thresholds), the system optimizes the balance between measurement precision and productivity. This parameter change enables the system to scale to larger datasets without linearly increasing processing time, thus maintaining productivity.