Automatic modeling method based on data element training set

The invention discloses an automatic modeling method based on a data element training set. A physical modeling process is simplified through the data element training set. The method comprises four core components of data element analysis, comparison, relation maintenance and modeling, and is suitable for processing various data types including spreadsheets and relational databases. The data element analysis component standardizes data input, and the comparison component ensures correct recognition of data elements through accurate and synonym matching. The relationship maintenance component allows the user to update the training set in response to unmatched elements. And the modeling component automatically constructs a physical model by using the verified data elements. According to the method, the modeling efficiency is remarkably improved, manual errors are reduced, and the method is particularly suitable for a large-scale data environment. 本发明公开了一种基于数据元训练集自动建模方法，通过数据元训练集简化物理建模流程。该方法包括数据元解析、比对、关系维护和建模四大核心