cd /news/natural-language-processing/toward-unsupervised-conceptual-metap… · home topics natural-language-processing article
[ARTICLE · art-57984] src=aclanthology.org ↗ pub= topic=natural-language-processing verified=true sentiment=· neutral

Toward Unsupervised Conceptual Metaphor Discovery: A Case Study in Online Immigration Discourse

Researchers at the University of Colorado Boulder introduced an unsupervised LLM-based method for discovering conceptual metaphors in political discourse, demonstrating its utility on online immigration discussions. The approach detects metaphorical expressions and clusters them into source domain categories without requiring predetermined inventories, enabling frame analysis of how different source domains frame the same target concept.

read2 min views2 publishedJul 13, 2026
Toward Unsupervised Conceptual Metaphor Discovery: A Case Study in Online Immigration Discourse
Image: Aclanthology (auto-discovered)
Abstract

In Conceptual Metaphor Theory (CMT), a metaphor is a systematic mapping from a concrete source domain (e.g., physical load) to a more abstract target domain (e.g., taxes), so that reasoning about concepts in the target domain is guided by inferences from the source domain. In this work, we propose that since different source domains can frame the same target in starkly different ways, the conceptual mappings evidenced by metaphorical expressions can guide computational political discourse analysis. We present a proof-of-concept for an unsupervised method that uncovers salient conceptual mappings from a corpus. Prior work in computational political metaphor analysis has drawn on CMT, but it typically requires a predetermined inventory of focused source and target domains. In contrast, we introduce a simple LLM-based method that detects metaphorical expressions from a corpus with strong performance, then clusters them to approximate source domain categories. We demonstrate its utility through a case study on online immigration discourse, showing that the resulting metaphor clusters provide context for frame analysis. We conclude by outlining future work needed to develop a robust framework for conceptual metaphor discovery in political discourse.- Anthology ID:

- 2026.nlpcss-1.11
- Volume:
[Proceedings of the Seventh Workshop on Natural Language Processing and Computational Social Science](/volumes/2026.nlpcss-1/)- Month:
  • July
  • Year:
  • 2026
  • Address:
  • San Diego
- Editors:
[Dallas Card](/people/dallas-card/),[Anjalie Field](/people/anjalie-field/),[Katherine Keith](/people/katherine-keith/),[Julia Mendelsohn](/people/julia-mendelsohn/)- Venues:
[NLP+CSS](/venues/nlpcss/)|[WS](/venues/ws/)- SIG:
- Publisher:
  • Association for Computational Linguistics
- Note:
- Pages:
  • 159–175
- Language:
- URL:
[https://aclanthology.org/2026.nlpcss-1.11/](https://aclanthology.org/2026.nlpcss-1.11/)- DOI:
[10.18653/v1/2026.nlpcss-1.11](https://doi.org/10.18653/v1/2026.nlpcss-1.11)- Cite (ACL):
[Toward Unsupervised Conceptual Metaphor Discovery: A Case Study in Online Immigration Discourse](https://aclanthology.org/2026.nlpcss-1.11/)(Leto & Pacheco, NLP+CSS 2026)- PDF:
[https://aclanthology.org/2026.nlpcss-1.11.pdf](https://aclanthology.org/2026.nlpcss-1.11.pdf)
── more in #natural-language-processing 4 stories · sorted by recency
── more on @university of colorado boulder 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/toward-unsupervised-…] indexed:0 read:2min 2026-07-13 ·