Specifications of the exact requirements for BIM-based rules are often much more complex than it appears from their direct language interpretation. It is because the complete rule interpretation involves knowledge from human experts. Therefore, detailed interpretation process of rules for the purpose of automated rule checking implementation must be able to capture such knowledge and it must be retained throughout the implementation process. In order to ensure both completeness and precision, the knowledge capture needs to be formalized. We propose a formalized way to capture these requirements using a knowledge representation approach. Detailed analysis of the structure of the rules points to the conceptual graph (CG) as a suitable method for this purpose due to its expressiveness, which allows unambiguous description of the requirements. Such unambiguous description can be understood by all the participants in the implementation efforts. The conceptual graph is a representation that conforms to the first order logic that makes it suitable for the job. Using the conceptual graph, rules can be broken into their atomic rules and incorporates the unwritten domain checking knowledge that if done right will remove ambiguities that often plague building codes. It also provides a standardized way to capture and document the model data requirements and the high level checking logic as their functional requirements. With this, a layer that often separates the rule experts and the implementers can be eliminated, resulting in clarity and immediate usefulness for the implementation. The validity of this assertion is demonstrated using the comparison to the actual corresponding implementation of the rules.
展开▼
