XML Table Index with Domain Indexes for Complex Data Querying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database systems face inefficiencies in accessing XML documents due to the need for prior knowledge of XML schema, slow processing of XPath-based queries, and inadequate handling of complex data types, leading to suboptimal query performance.
Innovation Solution
The implementation of a mechanism that uses a XML Table Index with virtual columns and domain indexes to store and index complex data types, allowing for efficient querying of XML documents without requiring prior schema knowledge and improving query speed by leveraging domain-specific properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If XML documents are stored in shredded form with base structures, then hierarchical relationships are captured, but prior knowledge of XML schema is required and extraneous data types are stored
Solution Approach 1:
The patent introduces an XML Table Index as an intermediary structure between the shredded XML documents and the query processing system. This index table contains only the columns relevant to XPath-based queries, acting as a mediator that filters out extraneous data types while preserving the necessary hierarchical relationships for efficient querying.
Solution Approach 2:
The patent extracts only the relevant columns from the full XML schema that are needed for XPath-based queries. By taking out only the necessary data types and structural relationships, the system avoids storing and processing extraneous information, thereby reducing complexity while maintaining query efficiency.
2Adaptability or versatility
If XML documents are stored in aggregate form as CLOB or BLOB, then no prior schema knowledge is required, but full scans are needed for XPath queries resulting in slow processing
Solution Approach 1:
The patent creates an XML Table Index in advance that pre-processes and organizes relevant columns from XML documents into a structured format. This preliminary action enables subsequent XPath queries to be executed efficiently on the pre-organized index rather than performing full scans on aggregate CLOB/BLOB storage.
Solution Approach 2:
The patent segments the XML document storage into two parts: the original aggregate CLOB/BLOB storage for complete documents, and a separate XML Table Index containing segmented, relevant columns for efficient querying. This segmentation allows the system to maintain schema independence while enabling fast XPath query processing on the indexed portion.
3Ease of operation
If full scan of all XML documents is performed for XPath queries, then all queries can be satisfied, but processing time is quite slow
Solution Approach 1:
The patent extracts relevant column data from XML documents into an XML Table Index that is optimized for XPath-based queries. By taking out only the necessary columns into this index structure, the system can satisfy XPath queries by accessing only the relevant indexed data rather than performing full scans of all XML documents.
4Productivity
If XML Table Index is used for XPath queries, then query speed is improved, but columns of complex data types are not handled efficiently
Solution Approach 1:
The patent extends the XML Table Index to handle multiple data types universally by integrating domain-specific indexes for complex data types such as text, image, audio, and video. This multi-functional approach allows the same XML Table Index structure to efficiently handle both simple and complex data types through appropriate domain-specific indexing mechanisms.
Solution Approach 2:
The patent creates a composite indexing structure that combines the XML Table Index with domain-specific indexes for complex data types. This composite approach integrates multiple indexing strategies into a unified system that can efficiently handle diverse data types including text, image, audio, and video columns within the same XML document collection.
Data Source
AI summary
Commonly searched nodes of complex data types contained in a collection of XML documents can be projected to, and their values can be stored in, corresponding columns in XML Table Index tables in a database system. In one embodiment, those columns stores logical pointers, rather than raw values, of the complex data types. Domain indexes are further provided to index at least one of the columns of complex data types in XML Table Indexes. The inclusion of complex data types and domain indexes provide a more efficient mechanism for searching domain specific data, such as audio, video, text stored in the collection of XML documents. Query optimizer logic of the database system can prepare one or more query execution plans for a submitted query including a plan that leverages XML Table Indexes and domain indexes and select a query execution plan based on cost information.