EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Mining Exile: An NLP-Based Sentiment and Theme Analysis of Iraqi Diaspora Poetry

Ismail Abdulwahhab Ismail, Salama A. Mostafa

Iraqi Literary and Cultural Review (ILCR) · 2025

DOI: 10.69513/ilcr.v3.i1.a3 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Iraqi Literary and Cultural Review (ILCR), 2025. Verfasst von Ismail Abdulwahhab Ismail und 1 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

This article applies computational methods—topic modelling and sentiment analysis—to a corpus of Iraqi diaspora poetry in Arabic and English in order to map recurrent themes and emotional patterns related to exile, nostalgia, loss, and belonging. It seeks to complement close readings of exilic poetry with corpus-based evidence that clarifies how Iraqi poets abroad collectively “write” exile in the late twentieth and early twenty-first centuries.

Methods

The study compiles a bilingual corpus of 1,200 poems by twelve Iraqi poets who have lived outside Iraq for significant periods since the 1990s. After standard NLP preprocessing (tokenisation, lemmatisation, and stop-word removal), Latent Dirichlet Allocation (LDA) topic modelling is used to identify thematic clusters, while lexicon-based and machine-learning sentiment analysis models are applied separately to Arabic and English subsets. Emotional categories such as sadness, joy, anger, fear, and hope are tracked at the line and poem level.

Results

The analysis reveals six dominant thematic groupings: (1) homeland and memory; (2) war and political violence; (3) routes of migration; (4) religion and ritual; (5) urban exile in Western cities; and (6) language, writing, and poetic self-reflexivity. Overall, the corpus exhibits strongly negative sentiment, with high proportions of sadness and fear, yet with pockets of positive affect tied to collective resistance, intimate relationships, and imagined return. Arabic poems show sharper oscillations between despair and hopeful defiance, while English-language texts by the same poets often adopt a cooler, more documentary tone.

Conclusions

The study demonstrates that computational approaches can effectively visualise shared affective and thematic patterns in Iraqi diaspora poetry, highlighting both convergence and divergence across languages. At the same time, it underscores the need for careful human interpretation of metaphor, irony, and historical context to avoid flattening the complexity of exilic experience into mere quantitative trends.

Zitieren

APA

Ismail, I. A., & Mostafa, S. A. (2025). Mining Exile: An NLP-Based Sentiment and Theme Analysis of Iraqi Diaspora Poetry. Iraqi Literary and Cultural Review (ILCR). https://doi.org/10.69513/ilcr.v3.i1.a3

Vancouver

Ismail IA, Mostafa SA. Mining Exile: An NLP-Based Sentiment and Theme Analysis of Iraqi Diaspora Poetry. Iraqi Literary and Cultural Review (ILCR). 2025. doi:10.69513/ilcr.v3.i1.a3

BibTeX
@article{Ismail2025,
  author = {Ismail, Ismail Abdulwahhab and Mostafa, Salama A.},
  title = {Mining Exile: An NLP-Based Sentiment and Theme Analysis of Iraqi Diaspora Poetry},
  journal = {Iraqi Literary and Cultural Review (ILCR)},
  year = {2025},
  doi = {10.69513/ilcr.v3.i1.a3}
}
Kostenlos starten →