Liuling Mo
Biographic Data
| ID | 7773341 |
|---|---|
| NAME | Liuling Mo |
| GIVEN NAMES | Liuling |
| FAMILY NAME | Mo |
| SIGNATURE | MO L |
| AFFILIATIONS | Institute of Psychology, Chinese Academy of Sciences |
| VERIFIED | No |
| TOTAL WORKS | 4 |
| TOTAL CITATIONS | 0 |
| WORKS AS AUTHOR | 4 |
| WORKS AS EDITOR | 0 |
| FIRST PUBLICATION YEAR | 2022 |
| LATEST PUBLICATION YEAR | 2025 |
| H-INDEX | 0 |
Suicidal ideation data augmentation and recognition technology based on large language models
摘要: 自杀已成为全球性公共卫生难题, 传统的自杀意念识别方法主要依赖患者主动求助, 而基于文本分析的自动识别模型则受限于标注数据的稀缺性。本研究创新性地提出一种基于大语言模型的数据增强方法, 旨在提升自杀意念文本识别的精度。研究采用双阶段设计:研究1聚焦于数据增强, 研究2验证增强效果。在研究1中, 选用ChatGLM3-6B和Qwen-7B-Chat作为底层模型, 结合有监督学习策略与零样本和少样本学习方法, 优化训练数据集质量。通过8组严谨的对比实验, 结果显示两类自研模型在数据增强方面表现卓越, 其处理后数据集的综合得分分别达到0.90和0.92, 显著优于基线模型(p < 0.001)。研究2进一步评估了数据增强对识别模型性能的影响, 结果表明, 增强后的模型在识别准确率和正确拒绝率指标上全面超越最佳基线模型(p < 0.001)。本研究不仅验证了基于大语言模型的数据增强方法在提升自杀意念识别模型性能方面的有效性, 还为心理健康领域的人工智能应用开辟了新方向。这种方法有望在保护…
The impact of pandemic mental cognition on cultural values
The study found that more collectivist regions are associated with a higher pathogen burden, and recognized the sense of uncertainty as its underlying cause. Results of this study validated and further developed the pathogen stress hypothesis in the context of the COVID-19 pandemic
Exploring the Suicide Mechanism Path of High-Suicide-Risk Adolescents—Based on Weibo Text Analysis
mental illness (especially depression) in high-suicide-risk adolescents is closely related to suicide stages, the later the suicide stage, the higher the diagnosis rate of mental illness. The suicide mechanism path in high-suicide-risk adolescents is: psychological pain→ hopelessness → suicide stages, indicating that psychological pain mainly affects suicide risk through hopelessness. Adolescents who are later in the suicide stages have fewer exp…
Transition from Depression to Suicidal Attempt in Young Adults
Our findings might contribute to the expansion of the interpersonal theory of suicide and have an impact on effective suicide prevention
No prominent works on this page.
Exploring the Suicide Mechanism Path of High-Suicide-Risk Adolescents—Based on Weibo Text Analysis
mental illness (especially depression) in high-suicide-risk adolescents is closely related to suicide stages, the later the suicide stage, the higher the diagnosis rate of mental illness. The suicide mechanism path in high-suicide-risk adolescents is: psychological pain→ hopelessness → suicide stages, indicating that psychological pain mainly affects suicide risk through hopelessness. Adolescents who are later in the suicide stages have fewer exp…
Transition from Depression to Suicidal Attempt in Young Adults
Our findings might contribute to the expansion of the interpersonal theory of suicide and have an impact on effective suicide prevention
The impact of pandemic mental cognition on cultural values
The study found that more collectivist regions are associated with a higher pathogen burden, and recognized the sense of uncertainty as its underlying cause. Results of this study validated and further developed the pathogen stress hypothesis in the context of the COVID-19 pandemic
Suicidal ideation data augmentation and recognition technology based on large language models
摘要: 自杀已成为全球性公共卫生难题, 传统的自杀意念识别方法主要依赖患者主动求助, 而基于文本分析的自动识别模型则受限于标注数据的稀缺性。本研究创新性地提出一种基于大语言模型的数据增强方法, 旨在提升自杀意念文本识别的精度。研究采用双阶段设计:研究1聚焦于数据增强, 研究2验证增强效果。在研究1中, 选用ChatGLM3-6B和Qwen-7B-Chat作为底层模型, 结合有监督学习策略与零样本和少样本学习方法, 优化训练数据集质量。通过8组严谨的对比实验, 结果显示两类自研模型在数据增强方面表现卓越, 其处理后数据集的综合得分分别达到0.90和0.92, 显著优于基线模型(p < 0.001)。研究2进一步评估了数据增强对识别模型性能的影响, 结果表明, 增强后的模型在识别准确率和正确拒绝率指标上全面超越最佳基线模型(p < 0.001)。本研究不仅验证了基于大语言模型的数据增强方法在提升自杀意念识别模型性能方面的有效性, 还为心理健康领域的人工智能应用开辟了新方向。这种方法有望在保护…
Medicine (4 works) · Psychology (4 works) · Medical emergency (3 works) · Poison control (3 works) · Suicide prevention (3 works) · Computer Science (2 works) · Psychiatry (2 works) · Social Psychology (2 works) · Social Psychology (2 works) · Suicidal ideation (2 works)