MOTIVATION: Medical text classification plays a critical role in clinical decision support, automated diagnosis, and biomedical research. However, deep learning models are highly susceptible to dataset-induced biases, such as label bias and keyword bias, which can lead to unreliable predictions and poor generalization in real-world clinical applications. Existing debiasing methods often either overcorrect informative samples or lack interpretability during inference, limiting their effectiveness in multilabel medical text classification tasks. RESULTS: We propose Cooperative Debiasing Network (CoDeNet), a cooperative training framework that mitigates dataset bias through dynamic sample reweighting and interpretable counterfactual inference. The framework consists of a primary classifier and a debias estimator, where the debias estimator quantifies sample-level bias and dynamically regulates the optimization process through an elastic scaling mechanism. In addition, a counterfactual postprocessing strategy explicitly isolates label-level and keyword-level biases to improve interpretability. Experiments conducted on the DepressionEMO and BDI-Sen datasets demonstrate that CoDeNet consistently improves classification performance over strong transformer-based baselines, including BERT and MentalBERT. In particular, CoDeNet achieves improvements of up to +6.57% macro-F1 on BDI-Sen and +2.16% macro-F1 on DepressionEMO, with especially strong gains on low-frequency clinical labels. The results indicate that CoDeNet effectively reduces dataset-induced bias while preserving model robustness and interpretability. AVAILABILITY AND IMPLEMENTATION: The source code and implementation details of CoDeNet will be publicly available on GitHub: https://github.com/66ccff39C5BB/CoDeNet.

Efficient code search techniques are crucial in accelerating software development by aiding developers in locating specific code snippets and understanding code functionalities. This study investigates code search methodologies, focusing on the emerging significance of semantic consistency in data augmentation techniques. While existing approaches predominantly enhance raw data, often requiring additional preprocessing and incurring higher training costs, this research introduces a pioneering method operating at the code and query representation levels. By bypassing the need for extensive data processing, this novel approach fosters an interactive alignment between code and query, augmenting the semantic coherence crucial for effective code search. An extensive empirical evaluation of a diverse dataset across multiple programming languages substantiates the efficacy of this approach in significantly enhancing code search model performance compared to traditional methodologies. The implementation is publicly available on GitHub, 1 offering an accessible resource for further exploration and application.

Code summarization is the task of automatically generating natural language descriptions from source code. Recently, pre-trained language models have gained significant popularity in code summarization due to their capacity to capture richer semantic representations of both code and natural language. Nonetheless, contemporary code summarization models grapple with two fundamental limitations. (1) Some tokens in the code are irrelevant to the natural language description and damage the alignment of the representation spaces for code and language. (2) Most approaches are based on the encoder-decoder framework, which is often plagued by the exposure bias problem, hampering the effectiveness of their decoding sampling strategies. To address the two challenges, we propose a novel pipeline framework named Reduce Redundancy then Rerank (Reˆ3). Specifically, a redundancy reduction component is introduced to eliminate redundant information in code representation space. Moreover, a re-ranking model is incorporated to select more suitable summary candidates, alleviating the exposure bias problem. The experimental results show the effectiveness of Reˆ3 over some state-of-the-art approaches across six different datasets from the CodeSearchNet benchmark.