Le Colloque > Résumés > Sayf Mohamed

Metaphors identification in large textual corpora has long been a challenge for computational linguistics and natural language processing. Metaphors are pervasive in human communication and cognition, but their often subtle and context-dependent nature makes automated detection difficult. Recent advances in artificial intelligence, particularly in deep learning and natural language understanding, have opened up promising new avenues for metaphor identification at scale.

Traditional computational approaches to metaphor identification relied heavily on hand-crafted rules, lexical resources, and statistical techniques. While these methods achieved some success, they were limited in their ability to capture the full complexity and variety of metaphorical language use. Modern AI techniques, especially those based on large language models and neural networks, offer a more flexible and context-aware approach to metaphor detection. The application of AI to metaphor identification in corpora has significant implications for linguistic research. It enables the analysis of metaphor use across much larger and more diverse datasets than was previously feasible, potentially revealing new insights into the cognitive processes underlying metaphor production and comprehension.

However, current AI tools still struggle with highly context-dependent or creative metaphors, and there are ongoing debates about how to define and annotate metaphors for machine learning purposes. In this article, we compare human metaphor identification to AI metaphor identification: This study compares metaphor identification performance between human annotators and AI systems using a corpus of tweets posted by Emmanuel Macron on the topic of climate change. The results reveal significant differences between human and AI metaphor identification. AI systems identified a higher overall number of potential metaphors, with the fine-tuned language model detecting 12% more metaphorical expressions than the human average. However, human annotation demonstrated greater proficiency in identifying novel or creative metaphors and showed higher sensitivity to contextual nuances. AI systems excelled in detecting conventional metaphors and domain-specific metaphors in technical areas. These divergent results highlight the complementary strengths of human and AI approaches while underscoring the challenges in automated metaphor detection, particularly in the unique linguistic environment of social media communications.

Bibliography

Adolphs, R. (2015). The unsolved problems of neuroscience. Trends in Cognitive Sciences, 19(4), 173-175.

Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., & Qin, Y. (2004). An integrated theory of the mind. Psychological Review, 111(4), 1036-1060.

Armstrong, S., & Sotala, K. (2015). How we're predicting AI–or failing to. In Beyond Artificial Intelligence (pp. 11-29). Springer.

Boden, M. A. (2016). AI: Its nature and future. Oxford University Press.

Bolukbasi, T., Chang, K. W., Zou, J. Y., Saligrama, V., & Kalai, A. T. (2016). Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. Advances in Neural Information Processing Systems, 29, 4349-4357.

Chen, Y., Elenee Argentinis, J., & Weber, G. (2016). IBM Watson: how cognitive computing can be applied to big data challenges in life sciences research. Clinical Therapeutics, 38(4), 688-701.

Kövecses, Z. (2010). Metaphor: A practical introduction. Oxford University Press.

Kovecses, Z. (2018) Metaphor in Media Language and Cognition: A Perspective from Conceptual Metaphor Theory. Lege Artis, 3(1): 124–141.

Lakoff, G., & Johnson, M. (1980). Metaphors we live by. University of Chicago Press.

Marcus, G. (2018). Deep learning: A critical appraisal. arXiv preprint arXiv:1801.00631.

Markoff, J. (2012, June 25). How many computers to identify a cat? 16,000. The New York Times.

Steen, G. (2010). A Method for Linguistic Metaphor Identification, John Benjamins Publishing Company, Amsterdam.

Thibodeau, P. H., & Boroditsky, L. (2011). Metaphors we think with: The role of metaphor in reasoning. PloS One, 6(2), e16782.

Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitive science and human experience. MIT Press.