Seminar | Ambiguity and Archive: Computational Hermeneutics of Conflict Poetry through RAG

Event organised by the Computational Humanities research group.

To register to the seminar, please fill in this form.

27 May 2025 – 4:30pm BST

Remote – Via Microsoft Teams.

In person – Details shared upon registration.

Jenny Kwok (University of Hong Kong), Ambiguity and Archive: Computational Hermeneutics of Conflict Poetry through RAG

Abstract

This presentation proposes a methodological bridge between computational literary studies and conflict historiography through AI-augmented archival analysis. Focusing on Northern Ireland’s Troubles poetry, the study leverages the Conflict and Politics in Northern Ireland Archive (CAIN) to construct a Retrieval-Augmented Generation (RAG) framework that dynamically contextualizes poetic ambiguity within historical narratives.

The framework reconciles the scalability of AI with humanities rigor by integrating close reading practices, machine-assisted contextualization, and archival metadata. It establishes a replicable model for analyzing contested histories while prioritizing political sensitivity through localized AI training, demonstrating how resource-limited institutions can conduct computationally intensive scholarship without dependence on proprietary systems.

A comparative analysis of humanistic and computational methods reveals that hybrid approaches—where archival grounding tempers machine learning outputs—reduce historical projection biases in sentiment analysis. This proves critical when interpreting poetic devices encoding sectarian dualism (e.g., metaphorized territoriality in Seamus Heaney’s work). The study further critiques the temporality of AI-archival integration, arguing that dynamic context-retrieval systems avoid flattening historical nuance compared to static training corpora.

The presentation concludes by proposing toolkits that enable scholars to employ for other conflict literatures, emphasizing adjustable parameters for geopolitical specificity. By decentralizing AI infrastructure and foregrounding archival multiplicity, this work advances interdisciplinary debates about computational criticism’s capacity to engage ethically with traumatic histories.

Bio

Dr. Jenny Kwok is Research Assistant Professor of the Faculty of Arts, University of Hong Kong, where she also serves as the Lab Coordinator of the Arts Technology Lab. Dr. Kwok’s research advances AI workflows for literary analysis, focusing on Irish conflict literature. She develops retrieval-augmented generation (RAG) systems to contextualize the ambiguity of Troubles-era poetry within historical archives, and fine-tunes LLMs for semantic analysis of Irish literary corpuses. Her methods prioritize sociopolitical sensitivity and literary nuances, countering AI’s tendency to flatten contested narratives.

Her forthcoming work proposes frameworks for democratizing AI in the humanities, emphasizing explainable AI (XAI) tools. This aligns with her reinterpretation of pre-digital methodologies (e.g., Josephine Miles’ concordance work) as blueprints for hybrid human-machine interpretation.

Dr. Kwok holds fellowship at the Cambridge Digital Humanities (2024-2025) and is Gale Scholar Asia Pacific, Digital Humanities Oxford (2026).

Leave a comment

Your email address will not be published. Required fields are marked *