基于领域对比自适应模型的大学生焦虑心理分析

doi:10.3778/j.issn.1673-9418.2311014

摘要/Abstract

摘要： 焦虑心理检测任务是利用自然语言处理技术对用户在社交媒体上发布的内容进行分析。为了促进对高校学生焦虑心理检测的研究，开发了一个微博数据集。针对数据集标注成本高等问题，从无监督的角度对突发公共卫生事件背景下的零样本数据进行分析，提出了基于领域对比学习的自适应模型（DAN-CL）。DAN-CL基于预训练语言模型，将源域文本和目标域文本输入到结合知识蒸馏的对抗网络中进行领域自适应训练，实现模型的知识迁移，并采用对比学习方法提高模型的泛化能力。基准实验结果表明，DAN-CL的性能优于现存的对比模型，消融实验进一步证明了模型中不同部分的有效性。对高校学生的心理状况进行分析，发现其心理焦虑变化情况与突发性公共卫生事件变化的趋势是一致的，受疾病发展与周围环境的影响其焦虑水平显著升高。对焦虑产生的原因进行具体分析并提出对策，为突发性公共卫生事件下以及后时代应对大学生的心理健康挑战提供理论支持。

关键词: 焦虑心理, 领域自适应, 对比学习, 情感分析

Abstract: The anxiety detection task involves the utilization of natural language processing techniques to analyze users?? content posted on social media. To promote research on anxiety detection among college students, a Weibo dataset has been developed. Addressing the high cost of data annotation, this paper adopts an unsupervised approach to analyze zero-shot data under the context of public health emergencies, and proposes an adaptive model called domain adaptive network based on contrastive learning (DAN-CL). DAN-CL, based on a pre-trained language model, integrates source domain text and target domain text into a knowledge distillation-based adversarial network for domain adaptive training, enabling knowledge transfer of the model. It also employs contrastive learning methods to enhance the generalization ability of the model. Benchmark experimental results show that DAN-CL outperforms existing comparative models, and ablation experiments further validate the effectiveness of different components within the model. Additionally, an analysis of the psychological status of college students reveals that changes in their anxiety levels align with the trends of sudden public health events. Under the influence of disease progression and the surrounding environment, their anxiety levels increase significantly. A specific analysis of the causes of anxiety and the countermeasures provides theoretical support for addressing college students?? mental health challenges during and after sudden public health events.

Key words: anxiety psychology, domain adaptive, contrastive learning, sentiment analysis

朱薇薇, 张益嘉, 刘贯通, 鲁明羽, 林鸿飞. 基于领域对比自适应模型的大学生焦虑心理分析[J]. 计算机科学与探索, 2024, 18(7): 1900-1910.

ZHU Weiwei, ZHANG Yijia, LIU Guantong, LU Mingyu, LIN Hongfei. Psychological Analysis of College Students?? Anxiety Based on Domain Comparison Adaptive Model[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(7): 1900-1910.

参考文献

[1] COVID-19 Mental Disorders Collaborators. Global prevalence and burden of depressive and anxiety disorders in 204 countries and territories in 2020 due to the COVID-19 pandemic[J]. Lancet, 2021, 398(10312): 1700-1712.
[2] WEISSKIRCH R S, CROSSMAN K A. The impact of death and dying education for undergraduate students during the COVID-19 pandemic[J]. Omega (Westport), 2022. DOI: 10. 1177/00302228221089818.
[3] WANG X M, HEGDE S, SON C, et al. Investigating mental health of US college students during the COVID-19 pandemic: cross-sectional survey study[J]. Journal of Medical Internet Research, 2020(9): e22817.
[4] TAQUET M, GEDDES J R, HUSAIN M, et al. 6-month neurological and psychiatric outcomes in 236?379 survivors of COVID-19: a retrospective cohort study using electronic health records[J]. Lancet Psychiatry, 2021, 8(5): 416-427.
[5] MCPHERSON A, MARTIN C R. A narrative review of the beck depression inventory (BDI) and implications for its use in an alcohol-dependent population[J]. Journal of Psychiatric and Mental Health Nursing, 2010, 17(1): 19-30.
[6] MCHUGH C M, CORDEROY A, RYAN C J, et al. Association between suicidal ideation and suicide: meta-analyses of odds ratios, sensitivity, specificity and positive predictive value[J]. BJPsych Open, 2019, 5(2): e18.
[7] 贾若男, 王晰巍, 李嘉兴.?突发公共卫生事件中多主体参与的社交网络群体聚类模型[J]. 情报理论与实践, 2022, 45(3): 28-34.
JIA R N, WANG X W, LI J X. A social network group clustering model for multi-subject participation in public health emergencies[J]. Information Studies: Theory and Application, 2022, 45(3): 28-34.
[8] CHANCELLOR S, BIRNBAUM M L, CAINE E D, et al. A taxonomy of ethical tensions in inferring mental health states from social media[C]//Proceedings of the 2019 ACM Conference on Fairness, Accountability, and Transparency, Atlanta, Jan 29-31, 2019. New York: ACM, 2019: 79-88.
[9] VAJRE V, NAYLOR M, KAMATH U, et al. PsychBERT: a mental health language model for social media mental health behavioral analysis[C]//Proceedings of the 2021 IEEE International Conference on Bioinformatics and Biomedicine, Dec 9-12, 2021. Piscataway: IEEE, 2021: 1077-1082.
[10] WANG X, ZHANG H J, CAO L, et al. Leverage social media for personalized stress detection[C]//Proceedings of the 28th ACM International Conference on Multimedia, Oct 12-16, 2020. New York: ACM, 2020: 2710-2718.
[11] TURCAN E, MURESAN S, MCKEOWN K. Emotion-infused models for explainable psychological stress detection[C]//Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Stroudsburg: ACL, 2021: 2895-2909.
[12] GHEISARI M, BAGHSHAH M S. Unsupervised domain adaptation via representation learning and adaptive classifier learning[J]. Neurocomputing, 2015, 165: 300-311.
[13] 孙道凯, 余如英, 吕乐, 等. 重大突发疫情危机背景下大学生焦虑心理及影响与保护性因素[J]. 西南师范大学学报(自然科学版), 2021, 46(10): 121-128.
SUN D Q, YU R Y, LV L, et al. On impact and protective factors of college students?? anxiety under background of major emergency epidemic crisis[J]. Journal of Southwest China Normal University (Natural Science Edition), 2021, 46(10): 121-128.
[14] SAID H, TAWFIK B S, MAKHLOURF M A. A deep learning approach for sentiment classification of COVID-19 vaccination tweets[J]. International Journal of Advanced Computer Science and Applications, 2023, 14(4): 530-538.
[15] RYU M, LEE G, LEE K. Knowledge distillation for BERT unsupervised domain adaptation[J]. Knowledge and Information System, 2022, 64: 3113-3128.
[16] SUN B, SAENKO K. Deep CORAL: correlation alignment for deep domain adaptation[C]//Proceedings of the 14th European Conference on Computer Vision, Amsterdam, Oct 8-16, 2016. Cham: Springer, 2016: 443-450.
[17] PENG X C, BAI Q X, XIA X D, et al. Moment matching for multi-source domain adaptation[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Seoul, Oct 27-Nov 2, 2019. Piscataway: IEEE, 2019: 1406-1415.
[18] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communication of the ACM, 2014, 63(11): 139-144.
[19] TZENG E, HOFFMAN J, SAENKO K, et al. Adversarial discriminative domain adaptation[C]//Proceedings of the 2017 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Honolulu, Jul 21-26, 2017. Piscataway: IEEE, 2017: 2962-2971.
[20] GOU J P, YU B, MAYBANK S J, et al. Knowledge distillation: a survey[J]. International Journal of Computer Vision, 2021, 129(6): 1789-1819.
[21] LIANG B, CHEN Z, GUI L, et al. Zero-shot stance detection via contrastive learning[C]//Proceedings of the 31st ACM Web Conference, Apr 25-29, 2022. New York: ACM, 2022:2738-2747.
[22] OORD A, LI Y, VINYALS O. Representation learning with contrastive predictive coding[EB/OL]. [2023-09-14]. https://arxiv.org/abs/1807.03748.
[23] WANG B, WANG S, CHENG Y, et al. InfoBERT: improving robustness of language models from an information theoretic perspective[C]//Proceedings of the 9th International Conference on Learning Representations, May 3-7, 2021.
[24] VASWANI A, SHAZEER N M, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, 2017. Red Hook: Curran Associates Inc., 2017: 5998-6008.
[25] AUGENSTEIN I, ROCKTASCHEL T, VLACHOS A, et al. Stance detection with bidirectional conditional encoding[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, Austin, Nov 1-5, 2016. Stroudsburg: ACL, 2016: 876-885.
[26] ALLAWAY E, SRIKANTH M, MCKEOWN K. Adversarial learning for zero-shot stance detection on social media[C]//Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Stroudsburg: ACL, 2021: 4756-4767.
[27] 张绮钰, 梁美贞, 范依丁, 等. 突发公共卫生事件下高校网络舆情引导策略研究[J]. 产业与科技论坛, 2022, 21(16): 287-288.
ZHANG Q Z, LIANG M Z, FAN Y D, et al. Research on the guidance strategy of university network public opinion under public health emergencies[J]. Industrial and Science Tribune, 2022, 21(16): 287-288.
[28] YEUNG N C Y, LAU J T F, YU N X, et al. Media exposure related to the 2008 Sichuan earthquake predicted probable PTSD among Chinese adolescents in Kunming, China: a longitudinal study[J]. Psychological Trauma: Theory, Research, Practice, and Policy, 2018, 10(2): 253-262．
[29] 闭翠珍, 李文, 李晓凤. 突发公共卫生事件背景下高校大学生心理健康问题研究——以贵州财经大学为例[J]. 西部学刊, 2021(16): 20-22.
BI C Z, LI W, LI X F. Research on the mental health problems of college students under public health emergencies: a case study of Guizhou University of Finance and Economics[J]. West Journal, 2021(16): 20-22.