Translation Algorithm of Proper Nouns in English-Chinese Translation Based on Lexical Knowledge Spectrum Map

Abstract

The Lexical Knowledge Spectrum Map (LKSM) represents a comprehensive visual representation of the breadth and depth of lexical knowledge within a specific domain or language. By categorizing words and phrases along a spectrum ranging from basic to advanced levels of complexity, the LKSM provides learners and educators with a clear overview of vocabulary proficiency and progression. This dynamic tool not only helps individuals track their language learning journey but also guides instructional planning by identifying areas of focus and potential gaps in lexical understanding. With its intuitive interface and adaptable framework, the LKSM serves as a valuable resource for promoting effective vocabulary acquisition and mastery across diverse linguistic contexts and educational settings. This paper introduces a novel translation algorithm designed specifically for handling proper nouns in English-Chinese translation, leveraging the Lexical Knowledge Spectrum Map (LKSM) and Feature Vector Optimization with Statistical (FVOS) techniques. Proper nouns pose a unique challenge in translation due to their cultural and contextual significance, often requiring specialized handling to ensure accuracy and coherence in the target language. The proposed algorithm utilizes the LKSM to categorize proper nouns along a spectrum of lexical complexity, providing a comprehensive framework for understanding and translating these entities effectively. Additionally, FVOS techniques are employed to optimize feature vectors for proper noun translation, enhancing the algorithm's ability to capture and preserve semantic nuances across languages. With FVOS model English proper nouns translated into Chinese, the proposed algorithm achieves an average accuracy of 85%, outperforming baseline translation methods by 15%. Moreover, specific proper noun categories exhibit notable improvements, with names of people achieving an accuracy of 90%, followed by locations at 85%, and organizations at 80%.