Machine Translation String Division Using Forward-Backward Likelihoods
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character string processing devices take a long time to process multiple strings, leading to potential delays in real-time translation.
Innovation Solution
A dividing device that calculates forward and backward division likelihoods for each word in a character string to determine optimal division points, allowing for faster chunk generation without processing multiple strings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple character strings are processed to ensure accurate division points, then division accuracy is improved, but processing time increases
Solution Approach 1:
The patent divides the character string into multiple candidate segments and evaluates each segment independently to determine division points. By processing the string in segmented portions rather than as a whole, the system maintains division accuracy while reducing the computational burden on the entire string, thereby decreasing processing time.
Solution Approach 2:
The patent calculates division likelihoods for only the necessary candidate segments rather than evaluating all possible division points in the entire character string. This partial action approach focuses computational resources on the most promising segments, ensuring accurate division point detection while minimizing unnecessary processing time.
2Adaptability or versatility
If multiple character strings are processed simultaneously for real-time translation, then translation coverage is improved, but processing speed decreases
Solution Approach 1:
The patent segments the character string into multiple candidate strings and processes them in parallel independent units. This segmentation allows the system to handle multiple translation candidates simultaneously without the processing time increasing linearly, as each segment can be evaluated independently and concurrently.
Solution Approach 2:
The patent performs preliminary division likelihood calculations for candidate segments before committing to full translation processing. By pre-evaluating which segments are most likely to contain division points, the system prepares translation candidates in advance, enabling faster real-time translation while maintaining comprehensive coverage.
Data Source
AI summary
A dividing device includes: an acquisition unit that acquires a plurality of words constituting a character string one by one from a head of the character string; a first calculation unit that calculates a forward division likelihood indicating a likelihood of dividing the character string at a position in front of a first word acquired by the acquisition unit and a backward division likelihood indicating a likelihood of dividing the character string at a position immediately after the first word; a detection unit that detects a division point based on the forward division likelihood and the backward division likelihood, the division point being a position at which the character string is divided; a generation unit that generates a chunk by dividing the character string at the division point; and an output unit that outputs the chunk.