Elamur Research⌕
☰
home / papers / 2026 / Detecting Experiential Intertextuality Across Migration Routes: Beyond Surface Similarity in French Narratives

Detecting Experiential Intertextuality Across Migration Routes: Beyond Surface Similarity in French Narratives

Summary

Whether they travel the Trans-Saharan corridor or the Balkan one, migrants on very different paths often report similar ordeals: police violence, exploitation by smugglers, risky crossings and separation from family. The authors define a new task, experiential intertextuality detection. It asks whether two sentences from different narratives describe the same kind of lived experience, even when they share little or no vocabulary. The corpus has 108 French-language interview narratives. From it, sentence pairs are generated and labelled automatically, with no manual annotation. Each pair is scored by five annotation-free method families: lexical overlap, sentence embeddings, POS-based structure, position-aware narrative features, and zero-shot LLM judgments. All scores are compared with 816 expert ratings. Surface and embedding measures align only weakly with the experts. A zero-shot LLM performs best among single methods, and a supervised combination of all features gives a clear further gain. Pairs from the departure phase of the journey show the strongest shared experience. The work shows that standard semantic similarity misses experiential parallels, and it offers HSS researchers a way to trace cross-route patterns in migration.

Key points

  • Corpus: 108 French narratives (99 Trans-Saharan, 9 Balkan) containing 5,922 unique sentences. Pairs are labelled with a 15-category theme lexicon (e.g., violence_police, passeur_exploitation). The data is not released for ethical reasons, but the code is public (IntertextMigra on GitHub).
  • Gold standard: 816 expert-scored pairs on a 0–1 scale. Inter-annotator agreement is low (Krippendorff's α = 0.27).
  • Single methods: Lexical baselines (TF-IDF, Jaccard, BM25), embeddings (CamemBERT, MiniLM, LaBSE, e5) and POS features all reach r ≤ 0.30. Qwen2.5-7B zero-shot is best at r = 0.38, about 56.6% of the noise ceiling.
  • Prompting effects: Few-shot examples make Qwen2.5-7B worse but substantially improve Mistral-7B.
  • Hybrid and narrative position: A supervised hybrid Ridge model over 31 features reaches r ≈ 0.45 (68.6% of the ceiling), a 21% gain over the best single method. Similarity in narrative position also correlates with expert scores (r = 0.131).

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.sigdial-1.10

@inproceedings{sari2026detecting,
  title = {{Detecting Experiential Intertextuality Across Migration Routes: Beyond Surface Similarity in French Narratives}},
  author = {Sari, Sakayo Toadoum and Robin, Nelly and Auzanneau, Michelle and Sais, Lakhdar and Petit, Véronique and Veniard, Marie and Jabbour, Said and Delorme, Fabien},
  booktitle = {Proceedings of the 27th Annual Meeting of the Special Interest Group on Discourse and Dialogue},
  year = {2026},
  doi = {10.18653/v1/2026.sigdial-1.10},
  url = {https://doi.org/10.18653/v1/2026.sigdial-1.10}
}

References

16 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Eneko Agirre, Carmen Banea, Daniel Cer, Mona Diab, Aitor Gonzalez-Agirre, Rada Mihalcea, German Rigau, and Janyce Wiebe. 2016. SemEval-2016 task 1: Semantic textual similarity, monolingual and cross-lingual evaluation. In Proceedings of SemEval, pages 497–511. ACL. find our card
  2. Lucie Bacon. 2022. La fabrique du parcours migratoire sur la route des Balkans: Co-construction des récits et écritures (carto)graphiques. Ph.D. thesis, Université de Poitiers. ↗
  3. David Bamman, Brendan O’Connor, and Noah A Smith. 2013. Learning latent personas of film characters. In Proceedings of ACL, pages 352–361. ACL. find our card
  4. Tommaso Caselli and Piek Vossen. 2017. The event StoryLine corpus: A new benchmark for causal and temporal relation extraction. In Proceedings of the Events and Stories in the News Workshop, pages 77– 86. ACL. find
  5. Daniel Cer, Mona Diab, Eneko Agirre, Iñigo Lopez-Gazpio, and Lucia Specia. 2017. SemEval-2017 task 1: Semantic textual similarity multilingual and crosslingual focused evaluation. In Proceedings of SemEval, pages 1–14. ACL. find
  6. Nathanael Chambers and Dan Jurafsky. 2008. Unsupervised learning of narrative event chains. In Proceedings of ACL, pages 789–797. ACL. find
  7. Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. Unsupervised cross-lingual representation learning at scale. In Proceedings of ACL, pages 8440–8451. ACL. find
  8. Agata Cybulska and Piek Vossen. 2014. Using a sledgehammer to crack a nut? lexical diversity and event coreference resolution. In Proceedings of LREC, pages 4545–4552. find our card
  9. Christopher Forstall, Walter Scheirer, David Bamman, and Gregory Crane. 2015. Modeling the scholars: Detecting intertextuality through enhanced word-level n-gram matching. Digital Scholarship in the Humanities, 30(4):503–515. ↗
  10. Fabrizio Gilardi, Meysam Alizadeh, and Maël Kubli. 2023. ChatGPT outperforms crowd workers for text-annotation tasks. Proceedings of the National Academy of Sciences, 120(30):e2305016120. find our card
  11. Andrew J Reagan, Lewis Mitchell, Dilan Kiley, Christopher M Danforth, and Peter Sheridan Dodds. 2016. The emotional arcs of stories are dominated by six basic shapes. EPJ Data Science, 5(1):1–12. find
  12. Muhammad Nihal Hussain, Kevin K Bandeli, Samer Al-khateeb, and Nitin Agarwal. 2018. Analyzing shift in narratives regarding migrants in Europe via blogosphere. In Proceedings of the International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction, pages 181–190. Springer. find
  13. Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence embeddings using siamese BERT-networks. In Proceedings of EMNLP-IJCNLP, pages 3982–3992. ACL. find
  14. David Ing, Fabien Delorme, Said Jabbour, Nelly Robin, and Lakhdar Sais. 2025. Text mining from migration narratives. In Proceedings of ECML-PKDD. find
  15. A Nelly Robin. 2014. Migrations, observatoire et droit: Complexité du système migratoire ouest-africain. In HdR. Université de Poitiers. find
  16. Zero-shot system prompt (French): Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, and 1 others. 2023. Mistral 7B. arXiv preprint arXiv:2310.06825. ↗

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/sari-2026-detecting-experiential-intertextuality-across-migration/