To address the phenomenon of similar classes, existing methods in few-shot continual relation extraction (FCRE) face two main challenges: non-representative prototypes and representation bias, especially when the number of available samples is limited. In our work, we propose Minion to address these challenges. Firstly, we leverage the General Orthogonal Frame (GOF) structure, based on the concept of Neural Collapse, to create robust class prototypes with clear separation, even between analogous classes. Secondly, we utilize label description representations as global class representatives within the fast-slow contrastive learning paradigm. These representations consistently encapsulate the essential attributes of each relation, acting as global information that helps mitigate overfitting and reduces representation bias caused by the limited local few-shot examples within a class. Extensive experiments on well-known FCRE benchmarks show that our method outperforms state-of-the-art approaches, demonstrating its effectiveness for advancing RE system.

Multi-label learning is a challenging computer vision task that requires assigning multiple categories to each image. However, fully annotating large-scale datasets is often impractical due to high costs and effort, motivating the study of learning from partially annotated data. In the extreme case of Single Positive Multi-Label Learning (SPML), each image is provided with only one positive label, while all other labels remain unannotated. Traditional SPML methods that treat missing labels as unknown or negative tend to yield inaccuracies and false negatives, and integrating various pseudo-labeling strategies can introduce additional noise. To address these challenges, we propose the Generalized Pseudo-Label Robust Loss (GPR Loss), a novel loss function that effectively learns from diverse pseudo-labels while mitigating noise. Complementing this, we introduce a simple yet effective Dynamic Augmented Multi-focus Pseudo-labeling (DAMP) technique. Together, these contributions form the Adaptive and Efficient Vision-Language Pseudo-Labeling (AEVLP) framework. Extensive experiments on four benchmark datasets demonstrate that our framework significantly advances multi-label classification, achieving state-of-the-art results.

Anh Duc Le, Nam Le Hai, Thanh Xuan Nguyen, Linh Ngo Van, Nguyen Thi Ngoc Diep, Sang Dinh, Thien Huu Nguyen. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Tung Nguyen, Tue Le, Hoang Tran Vuong, Quang Duc Nguyen, Duc Anh Nguyen, Linh Ngo Van, Sang Dinh, Thien Huu Nguyen. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Quang Duc Nguyen, Tung Nguyen, Duc Anh Nguyen, Linh Ngo Van, Sang Dinh, Thien Huu Nguyen. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

The development of large language models (LLMs) such as ChatGPT and Gemini has led to impressive advancements in question answering (QA) systems. However, they often rely on generic knowledge from the internet, resulting in hallucinated answers when applied to domain-specific QA tasks. Furthermore, their operational dependence on powerful GPUs poses challenges for practical software deployment. Building a QA systems for low-resource languages like Vietnamese is even more challenging due to the scarcity of labeled data and limited pre-trained language models. In this study, we aim at constructing a Vietnamese legal QA system using a retrieval-augmented generation approach to reduce incorrect outputs. Our focus is on improving answer generation accuracy by training small-scale LLMs suitable for real-world deployment. Our contributions are: (i) constructing Vietnamese legal provisions and QA datasets for training the system; and (ii) proposing methods to fine-tune language models with QA capabilities in the legal domain. Experimental results demonstrate that it is possible to train an LLM with fewer computational resources and a smaller dataset while maintaining effectiveness. Our findings highlight that designing an efficient training and fine-tuning strategy is crucial for overcoming these challenges, particularly in the context of Vietnamese legal question-answering tasks.