Welcome to Francis Academic Press

Frontiers in Medical Science Research, 2024, 6(11); doi: 10.25236/FMSR.2024.061105.

Research on the Application of Medical Text Matching Technology Combined with Twin Network and Knowledge Distillation in Online Consultation

Author(s)

Jinzhu Yang

Corresponding Author:
Jinzhu Yang
Affiliation(s)

AI Research, Dyania Health Inc, Jersey City, New Jersey, 07310, United States

Abstract

Online consultation has become an important way for people to seek advice from doctors during the epidemic, but the lack of face-to-face communication and differences in patients' self-reported symptoms make it difficult for doctors to accurately assess the condition. In this paper, a Chinese medical text matching model combining twin network and knowledge distillation technology is proposed, and a medical question answering system is established. The system uses the Siamese-UniLM model to capture deep text information, greatly improving the accuracy and F1 value, and reducing model parameters and computing resources through knowledge distillation technology to achieve efficient online reasoning. The experimental results show that the model significantly improves the system response speed and resource utilization efficiency, and provides a reliable solution for providing scientific and accurate medical advice.

Keywords

Medical Text Comparison, Twin Networks, Knowledge Distillation, Online Consultation

Cite This Paper

Jinzhu Yang. Research on the Application of Medical Text Matching Technology Combined with Twin Network and Knowledge Distillation in Online Consultation. Frontiers in Medical Science Research (2024), Vol. 6, Issue 11: 25-29. https://doi.org/10.25236/FMSR.2024.061105.

References

[1] Ishii S, Yamazaki T, Ito S, et al. Improving Object Coverage of Text-to-Image Generation by Object Matching. Proceedings of the Annual Conference of JSAI, 2022: 705. DOI: 10. 11517/pjsai. JSAI2022. 0_2O1GS705. 

[2] Chen L, Lan Y, Pang L, et al. Toward the Understanding of Deep Text Matching Models for Information Retrieval. J Intell Inf Syst, 2021. DOI: 10. 48550/arXiv. 2108. 07081. 

[3] Zhao X, Li Z, Wu S, et al. Deep Text Matching in Medical Question Answering System. IEEE Access, 2021, 5, 59-61. 

[4] Jang J S, Kong H Y. Question-Answering Pair Matching Based on Question Classification and Ensemble Sentence Embedding. Computer Systems Science and Engineering, 2023, 46 (9): 3471-3489. DOI: 10. 32604/csse. 2023. 035570. 

[5] Cao Y, Cao P, Chen H, Kochendorfer K. M, Trotter A. B, Galanter W. L, & Iyer R. K. Predicting ICU admissions for hospitalized COVID-19 patients with a factor graph-based model. Cham: Springer International Publishing. 2022, 6, 245-256

[6] Varatharajah Y, Chen H, Trotter A, & Iyer R. K. A Dynamic Human-in-the-loop Recommender System for Evidence-based Clinical Staging of COVID-19. In HealthRecSys@ RecSys, 2020, 8, 21-22