Machine Translation Analysis Result Selection Using Coincidence Matrix
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing analysis result selection apparatus and methods for example sentence driven machine translation systems have limitations in generating plural common word sequences, designating minimum word number, discretion of start position, and interchangeability between word and part of speech coincidence policies.
Innovation Solution
The proposed solution involves a cascade connection of units such as 'P.coincidence matrix generation unit', 'N.adjoining coincidence word detection element', and 'S.common word sequence detection unit' to generate plural common word sequences, and the introduction of 'D.dominant node detection unit' to compose temporal exclusive trees with arbitrary topmost nodes, allowing for part of speech coincidence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional coincidence detection is used, then single coincident word string is obtained, but limitation of beginning position and inability to generate plural common word sequences occurs
Solution Approach 1:
The detection process is divided into three sequential units: P.coincidence matrix generation unit creates a matrix of coincident positions, N.adjoining coincidence word detection element identifies adjoining coincident words, and S.common word sequence detection unit extracts common word sequences. This segmentation allows the system to handle multiple sequences with different start positions by processing coincidence information in stages, thereby generating plural common word sequences without requiring a single complex detection mechanism.
2Adaptability or versatility
If conventional exclusive tree generation is used, then only exclusive tree with topmost node identical to example tree is obtainable, but inability to generate exclusive trees with arbitrary topmost nodes occurs
Solution Approach 1:
The D.dominant node detection unit performs preliminary detection of the dominant node (the node representing the coincident word sequence) before composing the temporal exclusive tree. By identifying the dominant node in advance, the system can then compose the exclusive tree with this arbitrary node as the topmost node, rather than being constrained to use the example tree's topmost node. This preliminary action enables flexible tree generation while maintaining structural integrity.
3Reliability
If upper covering process is performed for forbidden tree detection, then comprehensive detection is achieved, but vast calculations are required
Solution Approach 1:
The system extracts and stores temporal exclusive trees that have already been determined to be forbidden (due to containing inappropriate structures or patterns) into a separate forbidden tree database. During analysis, instead of performing comprehensive upper covering processes to detect all forbidden trees, the system simply checks whether the generated exclusive tree exists in the pre-stored forbidden tree database. This extraction approach maintains high detection reliability while dramatically reducing computational burden by avoiding redundant calculations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This invention deals with improvement of example driven analysis part of filtering type which receives the OR tree which is a single tree expressing plural analysis trees obtained by application of grammar rules as a portion of input and exclusive trees which is the aggregation of temporal exclusive trees generated from example trees and exclusive trees stored in the system as another portion of input and composes single analysis tree. This invention has ability of obtaining plural common word sequences between the input sentence and example trees and ability of designating minimum common word number at the common word sequence and ability of decision of starting position of common word sequence at example trees and ability of replacing coincident words with coincident part of speech.