Information classification method, device and system based on Bayesian structure learning

Abstract

The invention discloses an information classification method, system and system based on Bayesian structure learning. Information is extracted according to materials provided by users, sorted into a structured data list, and learned through Bayesian structure according to historical user data sets and information databases. Obtain the co-occurrence relationship of each disease, construct a Bayesian network structure, aggregate the preceding and following diseases belonging to the same type in the Bayesian network to obtain the corresponding disease classification, and form a corresponding strategy plan to the user based on the expert experience knowledge map Output analysis results. Based on Bayesian structure learning, the present invention extracts information from user medical records and other materials and performs structural processing, combines historical data sets and information databases to obtain the association relationship of each disease, and combines expert experience knowledge maps to form corresponding reference strategy solutions. It can become an important reference information for doctors to assist in judgment and save processing time.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to an information classification method, device and system based on Bayesian structure learning. Background technique [0002] The classification and analysis of diseases is a necessary step in many scenarios. The current classification of diseases mainly relies on manual judgment, and the professional ability of professionals is uneven, which leads to confusion in the classification. The classification may be because the medical profession is not deep enough. For example, there are many causes of cancer. Doctors may have different expressions for thyroid cancer when issuing diagnosis certificates, such as goiter mass, thyroid malignant tumor, thyroid papillary malignant tumor, etc. These expressions are different. Corresponding to certain specific periods or types of thyroid cancer, but if the professionalism is not enough, it is easy to cause misclassification. Thi...

AI Technical Summary

Problems solved by technology

[0002] The classification and analysis of diseases is a necessary step in many scenarios. The current classification of diseases mainly relies on manual judgment, and the professional ability of professionals is uneven, which leads to confusion in the classification. The classification may be because the medical profession is not deep enough. For example, there are many causes of cancer. Doctors may have different expressions for thyroid cancer when issuing diagnosis certificates, such as goiter mass, thyroid malignant tumor, thyroid papillary malignant tumor, etc. These expressions are different. Corresponding to certain specific periods or types of thyroid cancer, but if the professionalism is not enough, it is easy to cause misclassification. This classification difficulty is more obvious in mental diseases. Therefore, it is necessary to use data-driven methods to assist in reasonable disease classification. Improve secondary reference reliability

Images