Accuracy and consistency of data are very important in database systems, particularly with the increasing use of sensitive data. This paper presents a semantic verification approach for database systems that checks the data integrity constraints using abstract interpretation on the octagon abstract domain. The approach abstracts the semantic meaning of SQL statements, enabling the analysis of the numerical relationship between data variables without focusing on syntactic analysis alone. The proposed system can automatically check if the integrity constraints specified by organizational policies are satisfied. The approach is tested on database benchmark applications and proved effective in detecting integrity constraint violations with high accuracy and scalability. In summary, this paper presents an efficient approach for checking integrity properties in complex data-driven systems.