Sino-Vietnamese Text Transcription using Word Embedding Approach

Sino-Vietnamese (aka Han Viet) characters, derived from ancient Chinese script, held prominence in Vietnam for centuries until the advent of the modern scripting system. Despite their historical significance, contemporary Vietnamese struggle to decipher these characters, hindering access to valuable cultural artifacts. This paper explores the complexities of Sino-Vietnamese characters, their divergence from modern Chinese characters, and the intricate process of converting them into readable texts. Notably, Sino-Vietnamese characters boast eight tones compared with Mandarin’s five, adding to their linguistic complexity. While remnants of these characters persist in artifacts and bibliographies, their comprehension requires linguistic expertise because of the multiple possible readings for a single character. Addressing this challenge requires the development of tools for automatic recognition and translation, enabling broader access to Vietnam’s cultural heritage. By facilitating the understanding of ancient texts, these efforts have contributed to the preservation and appreciation of Vietnamese linguistic and cultural traditions.