Article: English-derived place name recognition and translation based on knowledge graph and phonetic generation algorithm Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.27 pp.109 - 132 Abstract: This paper proposes a model for English-derived place name recognition and translation using a knowledge graph and a phonetic generation algorithm. By integrating multiple algorithms, the model enhances both recognition and translation accuracy. Experimental results show an AUC of 0.892. With 100 training iterations, the recognition error rate is 1.3%, the translation error rate is 0.8%, and the BLEU score reaches 67.3%, demonstrating strong performance. Practical analysis indicates the model has the lowest time consumption, minimal memory usage, superior classification performance, and over 95% fluency and consistency. The innovation of the research lies in the construction of a bidirectional dynamic interaction fusion mechanism through a knowledge graph, an LSTM algorithm, and a bidirectional matching maximum algorithm, targeting the semantic specificity of English-derived place names. This breaks the traditional one-way static fusion and achieves precise scene-based collaboration and closed-loop optimisation of semantics and phonetics. Inderscience Publishers - linking academia, business and industry through research

Title: English-derived place name recognition and translation based on knowledge graph and phonetic generation algorithm

Authors: Defeng Ma

Addresses: Teaching and Research Department of College English, Capital Normal University, Beijing, 100048, China

Abstract: This paper proposes a model for English-derived place name recognition and translation using a knowledge graph and a phonetic generation algorithm. By integrating multiple algorithms, the model enhances both recognition and translation accuracy. Experimental results show an AUC of 0.892. With 100 training iterations, the recognition error rate is 1.3%, the translation error rate is 0.8%, and the BLEU score reaches 67.3%, demonstrating strong performance. Practical analysis indicates the model has the lowest time consumption, minimal memory usage, superior classification performance, and over 95% fluency and consistency. The innovation of the research lies in the construction of a bidirectional dynamic interaction fusion mechanism through a knowledge graph, an LSTM algorithm, and a bidirectional matching maximum algorithm, targeting the semantic specificity of English-derived place names. This breaks the traditional one-way static fusion and achieves precise scene-based collaboration and closed-loop optimisation of semantics and phonetics.

Keywords: knowledge graph; phonetic generation algorithm; English-derived place names; recognition and translation; LSTM; bidirectional maximum matching algorithm.

DOI: 10.1504/IJICT.2026.152532

International Journal of Information and Communication Technology, 2026 Vol.27 No.27, pp.109 - 132

Received: 16 Oct 2025
Accepted: 16 Dec 2025
Published online: 25 Mar 2026 *

Title: English-derived place name recognition and translation based on knowledge graph and phonetic generation algorithm

Keep up-to-date