Smart Analysis Oriented Design of Health Statistics Data Model

Research Purpose: design a new data model to effectively analyze publicly opened health statistics data with fixed table structure. Proposed Methods: this paper established a data model based on star model, and used text similarity calculation method to identify the dimensions and correlations of indicators. Results: every table in the yearbook was split as per its properties and by indicator, time, geographic region and dimension, to form a star schema centered on "Yearbook data table", and through text similarity calculation method, the created "indicator table" and "dimension table" were utilized to determine the dimension of an indicator via custom macro program to find out its corresponding upper level indicators. Conclusion: the data model designed can effectively support the intelligent analysis of health statistic data and build foundation for health statistical data analysis systems.