A Depth Evidence Score Fusion Algorithm for Chinese Medical Intelligence Question Answering System

Area of Science:

• Medical Informatics
• Artificial Intelligence
• Natural Language Processing

Background:

• Medical question answering (QA) systems are crucial for efficient healthcare information retrieval.
• Traditional QA systems struggle with complex medical questions, limiting their diagnostic and treatment support capabilities.

Purpose of the Study:

• To develop an intelligent QA system for disease diagnosis and treatment in medical informationization.
• To propose a depth evidence score fusion algorithm to improve the accuracy of medical QA systems.

Main Methods:

• A novel text semantic evidence score based on Word2vec was developed to calculate semantic similarity.
• A depth evidence score fusion algorithm was proposed to measure text information through multiple algorithmic approaches.
• The algorithm was evaluated on a medical text corpus.

Main Results:

• The proposed depth evidence score fusion algorithm demonstrated superior performance in the evidence-scoring module of the intelligent QA system.
• The algorithm effectively measures text information and ensures accurate output of optimal candidate answers.
• Word2vec-based semantic evidence scores enhance the system's ability to handle complex medical queries.

Conclusions:

• The depth evidence score fusion algorithm significantly improves the accuracy and efficiency of Chinese Medical Intelligent Question Answering Systems.
• This approach offers a promising solution for intelligent disease diagnosis and treatment support.
• Further research can explore integrating this algorithm into broader medical informationization platforms.