Alias Mapping Table for Multi-Repository Attribute Search

The system pre-generates and stores inverted indexes for each attribute across all repositories before search operations. These indexes organize attribute values and their corresponding data object locations in advance, allowing the search module to quickly retrieve relevant data without performing full-text scans during actual search operations, thus reducing search time while maintaining completeness

The patent introduces an alias mapping table as an intermediary component between the search interface and the actual attribute repositories. This mapping table stores relationships between user-friendly alias names and actual attribute names across multiple repositories, enabling the search system to translate user queries into efficient multi-repository searches without requiring users to understand the underlying complex attribute structures, thereby improving search completeness while managing time consumption through optimized query routing

The system divides the global search space into multiple independent attribute-specific inverted indexes, one for each attribute across different repositories. Instead of maintaining a single large full-text index that consumes excessive memory, the search is segmented into attribute-level searches using pre-built indexes, significantly reducing peak memory consumption while ensuring all relevant data is captured

The patent creates lightweight copies of attribute metadata and indexing structures in the alias mapping table rather than duplicating entire data objects. The inverted indexes store only essential indexing information (attribute values and data object location pointers) rather than full data copies, enabling efficient searches across multiple repositories with reduced memory overhead