Morpheme Indexing for Semantic Text Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text retrieval technologies face challenges in quickly and accurately retrieving text data considering semantic attributes, with neighborhood retrieval methods being slow and prone to noise, and semantic structure analysis methods requiring high computational load and multiple passes.
Innovation Solution
An apparatus that generates an index including positions of morphemes and semantic attributes between them, allowing for efficient retrieval by matching query morphemes and attributes directly with text data morphemes and attributes, enabling one-pass semantic attribute-based retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neighborhood retrieval is used to retrieve text data, then retrieval accuracy is improved, but retrieval speed deteriorates and noise increases
Solution Approach 1:
The patent segments text data into morphemes (smallest meaningful units) and builds an index structure that organizes morphemes and their semantic attributes. This segmentation allows the retrieval system to quickly locate and compare only relevant morphemes rather than processing entire text passages, thereby improving retrieval speed while maintaining accuracy through precise morpheme-level matching.
2Measurement precision
If semantic structure analysis is used to retrieve text data, then retrieval accuracy is improved, but computational load increases and multiple passes are required
Solution Approach 1:
The patent performs preliminary semantic analysis by extracting semantic attributes from morphemes during the index building phase. Semantic attributes such as part-of-speech tags, dependency relationships, and meaning classifications are pre-computed and stored in the index structure. This preliminary action eliminates the need for complex semantic structure analysis during actual retrieval operations, reducing computational load and enabling single-pass retrieval while maintaining high accuracy.
3Speed
If morpheme-based indexing is implemented, then retrieval speed is improved, but index size increases
Solution Approach 1:
The patent merges multiple indexing dimensions (morpheme forms, semantic attributes, part-of-speech categories, and dependency relationships) into a unified index structure. By combining these elements into an integrated indexing system rather than maintaining separate indexes, the patent achieves efficient multi-criteria retrieval speed while avoiding the exponential growth in index size that would result from maintaining multiple independent indexes.
4Speed
If traditional word-based retrieval is used, then retrieval speed is maintained, but retrieval accuracy deteriorates due to expression variations
Solution Approach 1:
The patent changes the retrieval parameter from whole words to morphemes (smallest meaningful language units). By operating at the morpheme level rather than the word level, the system can match variations of expressions that share common morphemes. For example, different forms of the same root word can be matched through their shared morpheme base, improving retrieval accuracy while maintaining speed through efficient morpheme-level indexing and comparison.
Data Source
AI summary
An apparatus generates an index including positions of morphemes included in a target text data and semantic attributes between the morphemes corresponding to the positions. The apparatus gives information including positions of morphemes included in an input query and semantic attributes between the morphemes corresponding to the positions to the query, and executes a retrieval on the target text data, based on the information given to the query and the index.