GLGuangliang Liu

Guangliang Liu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Guangliang Liu has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Context-Aware Counterfactual Data Augmentation for Gender Bias Mitigation in Language Models

    by Shweta Parihar, Liu Guangliang, Natalie Parde, Lu Cheng

    A challenge in mitigating social bias in fine-tuned language models (LMs) is the potential reduction in language modeling capability, which can harm downstream performance. Counterfactual data augmentation (CDA), a widely used method for fine-tuning, highlights this issue by generating synthetic data that may align poorly with real-world distributions or creating overly simplistic counterfactuals that ignore the social context of altered sensitive attributes (e.g., gender) in the pretraining corpus. To address these limitations, we propose a simple yet effective context-augmented CDA method, Context-CDA, which uses large LMs to enhance the diversity and contextual relevance of the debiasing corpus. By minimizing discrepancies between the debiasing corpus and pretraining data through augmented context, this approach ensures better alignment, enhancing language modeling capability. We then employ uncertainty-based filtering to exclude generated counterfactuals considered low-quality by the target smaller LMs (i.e., LMs to be debiased), further improving the fine-tuning corpus quality. Experimental results on gender bias benchmarks demonstrate that Context-CDA effectively mitigates bias without sacrificing language modeling performance while offering insights into social biases by analyzing distribution shifts in next-token generation probabilities.

  • An Interpretable Data-Driven Medical Knowledge Discovery Pipeline Based on Artificial Intelligence

    by Shaobo Wang, Xinhui Du, Guangliang Liu, Hang Xing, Zengtao Jiao, Jun Yan, Youjun Liu, Haichen Lv, Yunlong Xia

    Difficulty in knowledge validation is a significant hindrance to knowledge discovery via data mining, especially automatic validation without artificial participation. In the field of medical research, medical knowledge discovery from electronic medical records is a common medical data mining method, but it is difficult to validate the discovered medical knowledge without the participation of medical experts. In this article, we propose a data-driven medical knowledge discovery closed-loop pipeline based on interpretable machine learning and deep learning; the components of the pipeline include Data Generator, Medical Knowledge Mining, Medical Knowledge Evaluation, and Medical Knowledge Application. In addition to completing the discovery of medical knowledge, the pipeline can also automatically validate the knowledge. We apply our pipeline's discovered medical knowledge to a traditional prognostic predictive model of heart failure in a real-world study, demonstrating that the incorporation of medical knowledge can effectively improve the performance of the traditional model. We also construct a scale model based on the discovered medical knowledge and demonstrate that it achieves good performance. To guarantee its medical effectiveness, every process of our pipeline involves the participation of medical experts.


  • Open Access

    Dynamic Augmentation Data Selection for Few-shot Text Classification

    by Guangliang Liu, Lifeng Jin, Owen Yuan, Jiayu Zhou

    Guangliang Liu, Lifeng Jin, Owen Yuan, Jiayu Zhou. Findings of the Association for Computational Linguistics: EMNLP 2022. 2022.