Research On The Knowledge Base Construction For Medical Insurance Auditing Based On Ontology

Computer-assisted audit has been gradually replacing the traditional manual auditing practices because of the increasing insured populations and the larger size of the fund these years. However, during the deployment of N-SAS,it takes a lot of manpower and time on corresponding data and analysing Policies, because different areas have their own policies and the information systems.To solve this problem, building a domain knowledge base is a good approach.This thesis takes medical insurance as an example, using the technologies in ontology learining field which contain words segmentation algorithm and words combination algorithm to acquire the important concepts of medical insurance auditing field and different relations between these concepts from dictionaries and medical insurance policies.Then, building an ontology for the medical insurance auditing which is based on the real situation of medical insurance field and the "Seven Steps" developed by Stanford .At last,the domain ontology is represented and stored by the protégétool to form a knowledge base for medical insurance auditing field.On this basis,after the comparative analysis of traditional concept similarity computation models based on the domain ontology,One of the models which takes the shortest distance between the concepts as the influencing factor was improved so that it will be much more applicable to the medical insurance auditing field.It can reduce the people's reliance and raise the efficiency when this model is used by auditors in the work of corresponding data in the early stages of the auditing work.