Tone classification of online medical services based on 1DCNN-BiLSTM

View article
PeerJ Computer Science

Main article text

 

Introduction

  1. Designed the ‘questionnaire on tone types of medical professional service providers’, which substantiates the rationale for categorizing the tones of doctors on online medical platforms into six distinct types.

  2. Created a proprietary dataset comprising platforms ‘Ding Xiang Doctor’ and ‘Chun Yu Doctor’, and trained it on a dual-channel model constructed with CNN and BiLSTM (1DCNN-BiLSTM), effectively recognizing six tones: normal, angry, stressed, tenderness, determination, and steadiness.

  3. Compared the performance of our model with other models based on evaluation metrics such as accuracy, precision, recall, F1 score, and Kappa value, and utilized ablation study to justify the appropriateness of our model’s settings and parameters.

Related work

Tone classification

Feature extraction

Proposed Methodology

Convolutional Neural Network

BiLSTM network

Proposed Model

Experiments and Results

Dataset

Parameter settings

Performance evaluation

Results

Classification performance

Confusion matrix

Ablation study

Conclusion

Supplemental Information

Questionnaire on tone types of medical professional service providers (English)

DOI: 10.7717/peerj-cs.2325/supp-1

Audio labeling task

DOI: 10.7717/peerj-cs.2325/supp-2

Questionnaire on tone types of medical professional service providers (Chinese)

DOI: 10.7717/peerj-cs.2325/supp-3

Detailed description and calculation method of features

DOI: 10.7717/peerj-cs.2325/supp-4

Executable code

To execute the audio code, use this in conjunction with the combined dataset at figshare: Huang, Cheng (2024). Raw Data .zip. figshare. Dataset. https://doi.org/10.6084/m9.figshare.25013849.v2.

DOI: 10.7717/peerj-cs.2325/supp-5

Additional Information and Declarations

Competing Interests

The authors declare that there are no competing interests.

Author Contributions

Cheng Huang conceived and designed the experiments, performed the experiments, analyzed the data, performed the computation work, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Peng Xie conceived and designed the experiments, performed the computation work, authored or reviewed drafts of the article, and approved the final draft.

Chunming Wu performed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Xiaojuan Liu analyzed the data, authored or reviewed drafts of the article, and approved the final draft.

Lin Zhang performed the computation work, authored or reviewed drafts of the article, and approved the final draft.

Ethics

The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):

The Academic Committee of the School of Journalism & Communication at Chongqing University has granted ethical approval for the conduct of this study.

Data Availability

The following information was supplied regarding data availability:

The executable code is available in the Supplementary Files.

The third-party data is available at:

- Dingxiang Doctor, https://dxy.com

- Chunyu Doctor, https://www.chunyuyisheng.com.

The combined dataset from the third party datasets are available at figshare: Huang, Cheng (2024). Raw Data.zip. figshare. Dataset. https://doi.org/10.6084/m9.figshare.25013849.v2.

Funding

This work was supported by the Scientific and Technological Research Program of Chongqing Municipal Education Commission (No. KJQN202301162), the Scientific Research Foundation of Chongqing University of Technology (No. 0121230235), the Chongqing Language and Writing Research Funds (No. yyk23208), and the Fundamental Research Funds for the Central Universities (No. 2023CDSKXYXW008). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

509 Visitors 515 Views 21 Downloads