Irregular building type checking is a significant part of building checking for seismic compliance, which is of immense significance to buildings’ seismic safety. An approach based on building information modeling (BIM) and ontology was proposed for identifying irregular building types, thus boosting efficiency and accuracy. Firstly, the rules for checking irregular building types were analyzed and then translated into semantic compliance checking rules that could be recognized by the computer. Meanwhile, the irregular building type checking ontology was created to represent the checking logic. Secondly, the building information model’s information to be checked was extracted using automatic parsing, such as floor opening area, and such information was extracted from the building finite element analysis result parameters by the pattern matching algorithm, such as torsional displacement ratio. The extracted information was organized on the basis of the proposed irregular building type checking ontology. Next, the sematic reasoning was utilized to obtain the checking results and generate the checking reports. Finally, a designed building was taken as an example to verify the feasibility and strong expansion ability of the proposed method, which could lay a technical foundation for further realizing the automatic building checking for seismic compliance.