Automated Content Translation via Key-Value Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional translation services are unable to handle content items with non-standard or user-defined tags, leading to corruption and inefficiency in translating content for websites with global presence, as they incorrectly translate user-defined tags and other non-intended information.

Innovation Solution

A system and method that identifies translatable and non-translatable components within content items, generates key-value pairs for translatable text, and uses a translation template to reconstruct translated content items, avoiding the translation of non-intended components by using a chain of processors and a translation service capable of handling pure text and limited predefined tags.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional translation services are used to translate content items with non-standard tags, then translation coverage is improved, but content integrity deteriorates due to incorrect translation of user-defined tags and non-intended information

Engineering Contradiction:

Improvetranslation coverageVSAvoidcontent integrity

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The content item is segmented into translatable text components and non-translatable structural components (user-defined tags, URLs, metadata). The system processes only the text content through translation services while preserving the structural integrity of non-text elements, thereby achieving both comprehensive translation coverage and content integrity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces an intermediary processing layer that sits between the content item and the translation service. This intermediary identifies and extracts only the translatable text portions, submits them for translation, and then reconstructs the content item by inserting translated text back into the original structure. This mediator prevents user-defined tags and non-text elements from being incorrectly translated.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If manual translation of content items is performed, then content integrity is maintained, but translation efficiency deteriorates due to time-consuming and expensive manual processing

Engineering Contradiction:

Improvecontent integrityVSAvoidtranslation efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system enables automated self-service translation by implementing intelligent text extraction and reconstruction capabilities. The system automatically identifies translatable content, processes it through translation services, and reconstructs the translated content items without requiring manual intervention, thereby achieving both high efficiency and maintained content integrity.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system replaces the manual mechanical translation process with an automated computational system. The automated system performs text extraction, translation submission, and content reconstruction operations that previously required human translators, dramatically improving translation efficiency while maintaining quality through structured processing.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If translation services process content items with user-defined tags, then translation completeness is improved, but translation accuracy deteriorates due to corruption of non-intended information

Engineering Contradiction:

Improvetranslation completenessVSAvoidtranslation accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system extracts only the pure text content from content items containing user-defined tags, URLs, and metadata. By separating translatable text from non-translatable elements before submission to translation services, the system achieves complete translation of all text portions while maintaining accuracy by preventing translation of non-text elements that should remain unchanged.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20230281397A1Content management systems for providing automated translation of content items

Publication Date: 2023.09.07 OPEN TEXT CORPORATION

AI summary

Systems, methods and products for automatic translation of content items, where in one embodiment a system includes one or more processors configured to identify components of a content item to be translated (translation items) and components not to be translated (non-translation items). The system generates a key-value pair corresponding to each translation item, where the key-value pair includes a key which comprises a unique identifier and a value which comprises a text string of the translation item. The system also generates a translation template that corresponds to the content item, wherein the translation template matches the content item except that the text string of each translation item is replaced by the key of the corresponding key-value pair. The system then provides a translation of each of the key-value pairs. The system generates a translated content item. By replacing each key in the translation template with the corresponding generated translation.