Fuzzy neural optimization for edge-assisted English terminology translation

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Accurate English terminology translation is essential for global communication, technical documentation, knowledge exchange, and domain-specific information access. However, semantic ambiguity, inappropriate terminology selection, and limited domain-specific accuracy can reduce the reliability of existing translation systems. This work proposes a Shuffled Frog Leaping Algorithm with Fuzzy Neural Networks (SFLA-FNN) for reliable and context-aware English terminology translation. An English Terminology Translation Accuracy Dataset containing 2,000 entries covering general and domain-specific vocabulary is used for evaluation. The data are processed through tokenization, stop-word removal, and Z-score normalization. The Fuzzy Neural Network (FNN) captures semantic relationships and uncertainty, while the Shuffled Frog Leaping Algorithm (SFLA) optimizes FNN parameters to improve convergence and translation performance. The proposed framework achieves 95.6% accuracy, 94.8% precision, 93.8% recall, and 94.2% F1-score, outperforming the evaluated baseline approaches under the same experimental setting. The findings demonstrate that combining fuzzy reasoning with evolutionary optimization can improve the accuracy, semantic consistency, and contextual reliability of terminology translation. The framework also provides a foundation for edge-assisted translation with low-latency inference, supporting technical and professional communication where terminology accuracy is important. Globally, the approach has potential to support more reliable cross-language knowledge exchange, technical communication, and domain-specific information access, while human expertise remains important for ambiguous or consequential translations.