EUVIMEDEUVIMED entdecken
Crossref

IPCL: Cross-Domain Invariant Prototypical Contrastive Learning for Few-Shot Text Classification

Shucheng Mao, Hua Cheng, Tao Cheng, Yiquan Fang, Mingyu Yang, Xiaoning Chen

International Journal of Pattern Recognition and Artificial Intelligence · 15.09.2026 · 40 (11)

DOI: 10.1142/s0218001426530010 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Pattern Recognition and Artificial Intelligence, online erschienen am 09.05.2026. Verfasst von Shucheng Mao und 5 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Prototypical networks are a typical method in few-shot learning. The instance-based training causes the model to learn specific domain features of the support set, leading to inadequate cross-domain generalization capabilities when faced with new domain data. To this end, this paper proposes a cross-domain Invariant Prototypical Contrastive Learning (IPCL) method. We extract invariant features (content features) and abstract invariant features (emotional features) from sample characteristics to construct invariant prototypes, which are domain-independent. Based on invariant representation learning, the model decouples latent features to obtain invariant features, which form the core basis for classification decisions. Abstract invariant features, refined from variable features (style features), reveal the underlying commonalities behind these variable features. Furthermore, we propose constructing auxiliary classification tasks that directly mine invariant prototype’s semantic features, combined with metric tasks to achieve cross-task information complementarity and enhance the model’s generalization ability to new domains. Experiments show that IPCL’s decoupling of latent variables is recognizable, achieving accuracies of 92.23% and 93.45% on the ARSC and SST-2 datasets, respectively. In cross-domain tasks, IPCL outperforms previous method by up to 1.11%, demonstrating IPCL’s superior cross-domain generalization capabilities.

Zitieren

APA

Mao, S., Cheng, H., Cheng, T., Fang, Y., Yang, M., & Chen, X. (2026). IPCL: Cross-Domain Invariant Prototypical Contrastive Learning for Few-Shot Text Classification. International Journal of Pattern Recognition and Artificial Intelligence, 40(11). https://doi.org/10.1142/s0218001426530010

Vancouver

Mao S, Cheng H, Cheng T, Fang Y, Yang M, Chen X. IPCL: Cross-Domain Invariant Prototypical Contrastive Learning for Few-Shot Text Classification. International Journal of Pattern Recognition and Artificial Intelligence. 2026;40(11). doi:10.1142/s0218001426530010

BibTeX
@article{Mao2026,
  author = {Mao, Shucheng and Cheng, Hua and Cheng, Tao and Fang, Yiquan and Yang, Mingyu and Chen, Xiaoning},
  title = {IPCL: Cross-Domain Invariant Prototypical Contrastive Learning for Few-Shot Text Classification},
  journal = {International Journal of Pattern Recognition and Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {11},
  doi = {10.1142/s0218001426530010}
}
Kostenlos starten →