Current Volume 8
Abstract
Correctly understanding and converting idiomatic language with cultural meanings remains a significant difficulty in AI and NLP practice. Native speakers rarely interpret idioms through direct meanings, as these language constructs depend entirely on social knowledge, personal and collective history, and surrounding details. AI translation systems have made plenty of progress recently because of the development of neural machine translation (NMT) and large language models (LLMS). However, these technologies still deal poorly with cultural language elements. This research checks the AI system's capability to translate culturally sensitive language expressions between various languages. Our research creates and assesses a collection of idioms and culturally specific phrases from English, Arabic, Chinese, French, and Swahili to examine different AI translation models, including Google Translate, Deepl, and GPT-based systems. At the same time, they translate these expressions into target languages. The assessment utilises automated tool scores (BLEU, METEOR, semantic similarity scoring) and human examiner assessments for faithfulness, fluency, and cultural appropriateness in translations. AI systems' translation process of idiom expressions requires a proposed flowchart demonstrating the steps from inputting idiomatic expressions through contextual disambiguation to generate target outputs. The table shows a comparative review that outlines how each algorithm functions and performs while translating idioms between various cultures. Transformer-based LLMs present better contextual understanding than previous statistical or rule-based approaches. Yet, they choose straightforward interpretations rather than implied meanings and generate cultural inaccuracies, mainly when working with languages involving minimal resources. The reported restrictions show that AI systems need to process culturally-enriched datasets and use inputs from linguistics with anthropology and cross-cultural study perspectives. This document advocates for fundamental changes in AI translation investigation by pushing AI systems beyond basic word-to-word translation. The research findings find crucial application in international communication, together with diplomatic practices, education systems, and content localization, because they ensure appropriate and respectful translation of cultural expressions.
Keywords
Idiomatic Translation, Culture-Specific Expressions, Natural Language Processing (NLP), Cross-Cultural AI, Machine Translation Models
Citations
IRE Journals:
Dilshat Azizov
"From Idioms to Algorithms: Translating Culture-Specific Expressions in AI Systems" Iconic Research And Engineering Journals Volume 7 Issue 10 2024 Page 543-551
IEEE:
Dilshat Azizov
"From Idioms to Algorithms: Translating Culture-Specific Expressions in AI Systems" Iconic Research And Engineering Journals, 7(10)
About IRE Journals
IRE Journals is an open access online journals established with an aim to publish high quality of research work in various diciplines. We have more than 6 years in publications. Over 2500+ papers already published.
Important Links
For Authors
Volume 7:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 6:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 5:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 4:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 3:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12