A Chaotic Neural Network Model for English Machine Translation Based on Big Data Analysis

Area of Science:

• Computational Linguistics
• Natural Language Processing
• Machine Translation

Background:

• Machine translation (MT) systems often produce errors in specialized academic texts.
• Informative academic texts present unique translation challenges due to complex sentence structures and specific terminology.

Purpose of the Study:

• To systematically categorize machine translation errors in computer science and technology abstracts.
• To propose effective post-translation editing strategies tailored to identified error types.
• To enhance the overall quality of machine translation in the computer science domain.

Main Methods:

• Utilizing a chaotic neural network model for big data analysis of English translations.
• Applying the Double Quantum Filter-Muttahida Quami Movement (DQF-MQM) error classification framework.
• Engaging translation professionals and computer experts for error confirmation and validation.

Main Results:

• Identified common machine translation error sources in academic computer science texts, including long/difficult sentences, passive voice, and terminology.
• Developed specific post-editing strategies to address these errors, focusing on maintaining source text logic and academic accuracy.
• Demonstrated that targeted editing improves the quality of machine-translated computer science abstracts.

Conclusions:

• The proposed post-editing strategies effectively mitigate common machine translation errors in computer science texts.
• Systematic error categorization and tailored editing are crucial for improving MT quality in specialized fields.
• This research offers valuable guidance for translators working with machine-translated computer science content.