Unified Database Index for Text Document Querying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database indexing methods are inefficient for handling complex queries in text document retrieval, such as searching for multiple words with a given distance or in specific locations, as they require multiple indexes and are costly in terms of processing time and storage space.
Innovation Solution
A database index structure that combines text and structure information, where each entry comprises a subset of words and the name of the hierarchical node, allowing for efficient queries by scanning for keys with the searched words and node name, and is ordered to optimize information retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple indexes are maintained in parallel for different query types, then query efficiency for specific query patterns is improved, but processing time and storage space increase significantly
Solution Approach 1:
The patent merges multiple separate indexes (word index, phrase index, location index) into a single unified index structure. Each index entry contains a key composed of multiple components (word, phrase, location) and a value containing document references. This consolidation allows the system to handle different query types using one index, eliminating the overhead of maintaining and searching multiple separate indexes while preserving query efficiency for all patterns.
Solution Approach 2:
The unified index structure is designed to serve multiple query functions simultaneously. The key composition (including word, phrase, and location elements) enables the same index to efficiently support simple word searches, consecutive word searches, and location-specific searches without requiring specialized indexes for each case.
2Productivity
If multiple indexes are maintained in parallel for different query types, then query efficiency for specific query patterns is improved, but storage space requirements increase
Solution Approach 1:
The patent merges multiple separate indexes (word index, phrase index, location index) into a single unified index structure. Each index entry contains a key composed of multiple components (word, phrase, location) and a value containing document references. This consolidation allows the system to handle different query types using one index, eliminating the overhead of maintaining and searching multiple separate indexes while preserving query efficiency for all patterns.
3Speed
If a simple word-based index is used, then single-word search efficiency is improved, but complex queries involving multiple words or locations become inefficient
Solution Approach 1:
The index key is segmented into multiple distinct components: word element, phrase element, and location element. This segmentation allows the system to efficiently handle simple word searches by matching only the word component, while also enabling complex queries by combining matches across multiple components. The segmented key structure provides flexibility for different query types without sacrificing performance in either case.
Solution Approach 2:
The patent extends the traditional one-dimensional word-based index by adding additional dimensions (phrase and location components to the key). This multi-dimensional key structure enables the index to capture not just which words appear in documents, but also their contextual relationships and positions, thereby supporting complex queries while maintaining the efficiency of simple word searches.
Data Source
AI summary
Certain example embodiments relate to a database index for indexing one or more text documents in a database. The text documents include one or more hierarchical nodes, and each node includes one or more words. The database index includes at least one entry, with each entry including a key. The key, in turn, includes a subset of words occurring in one of the hierarchical nodes of the text documents and the name of the respective hierarchical node. Associated with each key is a value including one or more references to the text documents in which the subset of words occurs.