Dutch Metaphor Extraction from Cancer Patients'Interviews and Forum Data using LLMs and Human in the Loop

2025-11-09

AI Summary

Metaphor identification in Dutch-language interviews and online forum texts from cancer patients remains challenging due to linguistic complexity and scarce annotated resources. Method: This study proposes a human-in-the-loop metaphor extraction framework leveraging large language models (LLMs), integrating chain-of-thought reasoning, few-shot learning, and self-prompting techniques to enhance metaphor detection performance in low-resource language settings; human-LLM collaborative verification ensures high annotation fidelity. Contribution/Results: We introduce HealthQuote.NL—the first publicly available, high-quality Dutch corpus of cancer-related metaphors—alongside reusable prompt templates and a standardized processing pipeline. Empirical evaluation demonstrates the effectiveness and robustness of LLMs for clinical metaphor identification across diverse discourse genres. The framework advances patient-centered communication, supports shared decision-making, strengthens health literacy, and provides a linguistic foundation for personalized care pathway design.

Technology Category

• Natural Language Processing: Language Grounding & Multi-modal NLP
• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Analogy

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Metaphors and metaphorical language (MLs) play an important role in healthcare communication between clinicians, patients, and patients'family members. In this work, we focus on Dutch language data from cancer patients. We extract metaphors used by patients using two data sources: (1) cancer patient storytelling interview data and (2) online forum data, including patients'posts, comments, and questions to professionals. We investigate how current state-of-the-art large language models (LLMs) perform on this task by exploring different prompting strategies such as chain of thought reasoning, few-shot learning, and self-prompting. With a human-in-the-loop setup, we verify the extracted metaphors and compile the outputs into a corpus named HealthQuote.NL. We believe the extracted metaphors can support better patient care, for example shared decision making, improved communication between patients and clinicians, and enhanced patient health literacy. They can also inform the design of personalized care pathways. We share prompts and related resources at https://github.com/aaronlifenghan/HealthQuote.NL

Problem

• Extracting Dutch metaphors from cancer patient interviews and online forums
• Evaluating LLM performance on metaphor detection using various prompting strategies
• Creating verified metaphor corpus to improve healthcare communication and care

Innovation

• Extracted Dutch metaphors using large language models
• Applied chain of thought and few-shot prompting strategies
• Verified metaphors through human-in-the-loop validation process