Jul 2026· Iraqi Literary and Cultural Review (ILCR)· 0 citations· 12 references
TL;DR
The paper argues that NLP should operate as an interpretive assistant rather than an autonomous literary translator in translating Iraqi poetry into English, and proposes a culturally aware, human-in-the-loop framework for supporting literary translation.
Abstract
ABSTRACT
Objectives: This article critically examines the role of natural language processing (NLP) in translating Iraqi poetry into English, focusing on its usefulness and its risks. It addresses three linked issues: computational readings of poetic imagery, translation of Iraqi dialectal and cultural markers, and ethical questions raised when automated systems interpret poems shaped by historical memory, place, rhythm, and affect.
Methods: The study proposes a culturally aware, human-in-the-loop framework for supporting literary translation. The framework combines corpus preparation, dialect-sensitive annotation, metaphor tracking, named-entity recognition, sentiment and stance analysis, and translator-centered evaluation. The method is illustrated through constructed and paraphrased examples, allowing the framework to be tested later on a verified corpus of Iraqi poems.
Results: The analysis indicates that NLP can assist translators by identifying recurrent images, place names, semantic fields, repetition patterns, lexical density, and dialectal signals. These outputs can provide useful linguistic evidence for interpretation. However, standard NLP and machine translation systems often flatten metaphor, misread local references, normalize dialect, ignore prosody, and produce fluent English renderings that reduce cultural resonance. Such distortions are especially serious when Iraqi poetry carries memory of war, displacement, loss, and belonging.
Conclusions: The paper argues that NLP should operate as an interpretive assistant rather than an autonomous literary translator. A responsible approach to Iraqi poetry requires a hybrid workflow in which computational tools generate evidence, human translators interpret it with cultural sensitivity, and culturally informed evaluation protects ambiguity, voice, historical specificity, and poetic meaning without replacing the translator's critical and responsible ethical judgement.
ABSTRACT
Objectives: This article examines how natural language processing can support the study, translation, preservation, and interpretation of Iraqi cultural texts. It argues that NLP becomes most valuable for the Iraqi humanities when computational methods are combined with cultural knowledge, philological care, translation ethics, and close literary interpretation.
Methods: The study adopts a qualitative and conceptual methodology informed by Arabic NLP, digital humanities, corpus linguistics, translation studies, and heritage informatics. It proposes an applied framework for building Iraqi cultural corpora, processing multilingual and historically layered texts, supporting human-in-the-loop translation, and analyzing literary and cultural patterns across large collections. The framework emphasizes corpus curation, metadata design, linguistic annotation, dialect sensitivity, and collaboration between computational specialists and humanities scholars.
Results: The analysis shows that NLP can improve access to Iraqi cultural materials through OCR, text normalization, named-entity recognition, topic modeling, terminology extraction, corpus alignment, machine-assisted translation, and computational literary analysis. These tools can help identify recurrent themes, places, figures, genres, and linguistic patterns. However, the article also identifies major challenges, including dialectal variation, script inconsistency, low-resource languages, cultural ambiguity, weak metadata, fragmented archives, and the risk of decontextualized algorithmic reading.
Conclusions: NLP should be treated not as a replacement for humanistic scholarship, but as a computational extension of it. For Iraqi cultural texts, sustainable progress requires curated corpora, multilingual data policies, ethical governance, translator review, heritage-sensitive metadata, and cooperation among universities, museums, libraries, archives, and local communities, while preserving interpretive depth, textual specificity, and cultural responsibility in future research infrastructures across Iraq today.
Tetyana Panyok· Iraqi Literary and Cultural...· 0 citations
The results show that LLMs and Google Translate consistently outperform specialized MT systems in terms of fluency, meaning preservation, and lexical-thematic alignment.
Beatriz Ribeiro Borges, P. H. R. Gabriel, E. Faria· International Journal of Dat...· 0 citations
Large language models (LLMs) have rapidly expanded the horizon of AI-assisted literary translation, yet the mechanisms of how prompts construct translation purpose are theoretically underexplored. This study, based on Vermeer’s Skopos Theory, conceptualizes prompts as a “digital translation brief” and examines the systematic regulation of four function-oriented prompts: baseline (P0), reader-oriented (P1), form-oriented (P2), and culture-oriented (P3) in DeepSeek’s translation strategies for English-to-Indonesian poetry translation, with Emily Dickinson’s Hope is the Thing with Feathers as the source text. The study finds, through comparative close reading across the four prompt conditions, as well as against two published human translations, that each prompt orientation produces distinct and observable shifts in diction, imagery construction and formal expression: P1 sacrifices collocational stability for literary register and emotional intensity at the expense of; P2 replicates structure but risks over-compliance which undermines target-reader adequacy; P3 allows discourse-level metaphorical coordination and selective dependence on the source-text. Human translators’ choices emerge from cultural memory, aesthetic intentionality, and poetic agency. DeepSeek’s functional adaptability is largely a matter of probabilistic generation and prompt compliance. The findings re-conceptualize prompts not only as technical input instructions, but as purposive regulators of translation, offering a translation-theoretic framework for analysing LLM behaviour and practical guidance for designing Skopos-informed prompts for AI-assisted literary translation.
Haijin Li· Journal of Translation and L...· 0 citations
Poetry is a unique form of expression valued for its role in preserving cultural heritage. Analyzing Arabic poetry is time-consuming and requires a high level of linguistic expertise; therefore, computational methods are useful, as they enable large-scale, extensive, and systematic analysis of poetry, thereby improving its accessibility for researchers and students. This article presents the first systematic review of natural language processing (NLP) and machine learning (ML) approaches for Arabic poetry. It addresses the question of which research tasks, methodologies, datasets, and evaluation approaches have been applied to Arabic poetry, and which trends and research gaps can be identified in the existing literature. In accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), the author conducted an exhaustive search across six major academic databases (ACL Anthology, IEEE Xplore, ACM, SpringerLink, Science Direct, and Google Scholar) for relevant studies published between January 2010 and May 2025. Eligibility was evaluated in several phases, and re-examination was conducted to ensure accuracy. The author performed task-level categorization, extracted key characteristics from each study, synthesized the findings, and presented them in tables and figures to highlight the main trends and research gaps in the literature. This study presents the first structured task-level synthesis of the field, identifying methodological trends, detecting evaluation inconsistencies, and highlighting research gaps that have not been critically consolidated before. Furthermore, the author assembled a comprehensive collection of available datasets and resources to promote standardized assessment.
Fan tweets vividly reflect fan register, a distinctive variety of digital language marked by creative and culturally embedded linguistic practices. Among its most salient features in Thai is the use of English loanwords and transliterations. The present study investigates 172 instances of English loanwords and transliterations drawn from two datasets of Thai fan tweets (June 2024-May 2025 and April-May 2026), examining their forms, functions, and rendering in English by automatic translation tool of Twitter (X). Five forms are identified: direct loanwords, semantic extensions, transliterations, hybrid compounds, and clippings and abbreviations, serving five functions: lexical gap filling, in-group identity marking, expressive function, economy of expression, and prestige and modernism. The study further examines how these items are rendered in English by automatic translation, identifying four translation strategies, namely, back-translation, naturalisation, substitution, and deletion, alongside two mechanisms underlying mistranslation: orthographic ambiguity and semantic misinterpretation. The findings demonstrate that while automatic translation can facilitate interlingual and intercultural fan communication, it continues to struggle with the orthographic instability, fan-specific meanings, and pragmatically embedded language use that characterise English loanwords and transliterations in Thai fan tweets, posing barriers to pragmatic equivalence and highlighting the need for developers to address variant spellings, community-specific semantics, and pragmatic context.
Nicha Klinkajorn· LEARN Journal : Language Edu...· 0 citations
This article analyzes linguistic and stylistic problems that arise in translation processes in the age of artificial intelligence, particularly through neural machine translation and large language models. The study acknowledges the advantages of AI-assisted translation, including speed, multilingual processing and the production of preliminary drafts. However, the main focus is placed on semantic ambiguity, contextual misinterpretation, literal rendering of idiomatic expressions, inconsistency in terminology, stylistic neutralization and the loss of cultural connotations. The article proposes an analytical model for assessing translation quality based on linguistic, stylistic and pragmatic criteria. The findings indicate that artificial intelligence can function as an effective auxiliary tool in translation, but human translators and post-editing remain essential in literary, academic, legal, medical and journalistic texts
Maftuna Choriyeva· Journal of Universal Science...· 0 citations