Psychological Education Health Assessment Problems Based on Improved Constructive Neural Network

Area of Science:

• Computational linguistics
• Mental health informatics
• Artificial intelligence in healthcare

Background:

• Traditional mental health assessment methods using word frequency lack contextual understanding.
• Lexicon sparsity and small lexicon size limit the effectiveness of traditional methods like Linguistic Inquiry and Word Count (LIWC).

Purpose of the Study:

• To propose and evaluate a Convolutional Neural Network (CNN)-based method for mental health assessment.
• To leverage CNN's contextual semantic extraction capabilities to overcome limitations of traditional approaches.
• To improve the accuracy and reliability of mental health status assessment using online text data.

Main Methods:

• Utilized online text data for mental health assessment.
• Developed a mental health assessment model based on Convolutional Neural Network (CNN).
• Evaluated the CNN model using CLPsych2017 measurement indicators, comparing it with FastText and CNN + Word2Vec.

Main Results:

• The CNN-based method outperformed traditional approaches across all evaluation indicators.
• Achieved superior F1 score (0.51) and Accuracy (ACC) (0.69) with the CNN model.
• CNN + Word2Vec achieved ACC of 0.67 and F1 of 0.49, indicating strong performance.

Conclusions:

• The proposed CNN-based method is feasible and effective for mental health assessment.
• CNN models demonstrate significant advantages in contextual semantic extraction for mental health analysis.
• This approach offers a promising advancement in utilizing digital data for mental health evaluation.