Automated Grammatical Error Correction via Round-Trip Translation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manual detection and correction of grammatical errors in text sequences by non-native English speakers are time-consuming and inefficient, particularly for those with limited English proficiency.

Innovation Solution

A computer-implemented method using round-trip translations across multiple languages to generate candidate text sequences with alternative grammatical options, scored for grammaticality, to automatically correct grammatical errors in a text sequence.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual detection and correction of grammatical errors is performed, then accuracy of correction can be maintained, but time consumption increases significantly

Engineering Contradiction:

Improveaccuracy of grammatical error correctionVSAvoidtime consumption for error detection and correction

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary actions by translating the source text into multiple target languages and then back-translating them before final correction. This round-trip translation process prepares multiple candidate corrections in advance, which are then evaluated against the original text to identify grammatical errors and generate corrected versions, thereby reducing the time needed for manual proofreading while maintaining accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces intermediary elements (multiple target languages and back-translation processes) as mediators between the source text and the final corrected output. By translating through intermediate languages and comparing back-translations with the original, the system automatically identifies grammatical errors and generates corrections, eliminating the need for time-consuming manual review while preserving correction accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If round-trip translation through multiple languages is performed, then grammatical error detection accuracy improves, but processing complexity increases

Engineering Contradiction:

Improvegrammatical error detection accuracyVSAvoidprocessing system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The processing system performs multiple functions using a unified approach: it translates source text to multiple target languages, back-translates each target language to the source language, compares back-translations with the original text, identifies grammatical errors, and generates corrected versions. This multi-functional system handles diverse language pairs and error types through a single automated framework, managing complexity while improving detection accuracy

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system creates copies of the source text in multiple target languages and then creates back-translated copies returning to the source language. By generating and comparing multiple copies (original source text, multiple target translations, and multiple back-translations), the system automatically identifies grammatical errors through comparison, improving detection accuracy while the automated copying process manages the complexity of handling multiple language versions

Inventive Principle:

Principle #26Copying

3Productivity

If automated correction system is implemented, then productivity increases, but handling of complex and compound errors may be insufficient

Engineering Contradiction:

Improvespeed of error correctionVSAvoidability to handle complex and compound errors

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system segments the grammatical error correction process into distinct steps: translation to multiple target languages, back-translation to source language, comparison of back-translations with original text, identification of grammatical errors, and generation of corrected versions. This segmented approach allows the automated system to handle each aspect separately, improving productivity while maintaining the ability to detect complex and compound errors through systematic comparison across multiple language representations

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9342499B2Round-trip translation for automated grammatical error correction

Publication Date: 2016.05.17 EDUCATIONAL TESTING SERVICE

AI summary

Systems and methods are provided for correcting a grammatical error in a text sequence. A first text sequence in a first language is received. The first text sequence is translated to a second language to provide a first translated text. The first text sequence is translated to a third language to provide a second translated text. The third language is different from the second language. The first translated text is translated to the first language to provide a first back translation. The second translated text is translated to the first language to provide a second back translation. A plurality of candidate text sequences that include features of the first back translation and the second back translation are determined. The plurality of candidate text sequences include alternative grammatical options for the first text sequence. The plurality of candidate text sequences are scored with the processing system.