Machine Translation Device Using Intermediate Annotations

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional multilingual translation systems using a natural language as an intermediate language face issues such as data loss and reduced translation accuracy due to the inability to retain sufficient data, leading to mistranslations and loss of nuances between source and target languages.

Innovation Solution

A machine translation system employing two translation engines, where the first engine translates text into an intermediate language with added annotations, and the second engine uses these annotations to translate into the target language, utilizing dictionaries and linguistic annotation data to preserve source language data and nuances.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a natural language is used as an intermediate language in multilingual translation, then translation between arbitrary languages can be achieved, but data loss occurs and translation accuracy deteriorates

Engineering Contradiction:

Improvetranslation coverageVSAvoiddata loss

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The translation process is segmented into two distinct stages: first translating from source language to intermediate language while preserving source language data, then translating from intermediate language to target language using preserved data. This segmentation allows each stage to focus on specific translation needs without losing information.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Source language data is preserved and stored before the second translation stage begins. This preliminary preservation of data ensures that when translating from intermediate language to target language, the original source language information is still available to maintain translation accuracy and prevent data loss.

Inventive Principle:

Principle #10Preliminary action

2Device complexity

If source language expressions are translated into one expression in intermediate language, then translation simplification occurs, but nuance information is lost

Engineering Contradiction:

Improvetranslation process complexityVSAvoidnuance information

Core Design Contradiction:

Device complexityVSLoss of information

Solution Approach 1:

Multiple expression options and nuance information are preserved from the source language before translation to the intermediate language. This preliminary preservation ensures that even when expressions are simplified in the intermediate language, the original nuance information remains available for the final translation to the target language.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9002695B2Machine translation device, method of processing data, and program

Publication Date: 2015.04.07 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Provided are two types of translation engines. One of the translation engines translates a text written in a source language X into an intermediate language C, and the other translates the intermediate language C into a target language Y. The X-C translation engine translates an original text written in the source language X into a text in a predetermined intermediate language C as well as acquires and adds annotations obtained by this translation to the translated text written in the intermediate language C. The C-Y translation engine references annotations, which are obtained by the X-C translation engine and added to the text in the intermediate language C, and translates the text in the intermediate language into a text in the target language Y sentence.