Search

Semantics of Business Vocabulary and Rules (SBVR) is a standard that is applied in describing business knowledge in the form of controlled natural language. Business process designers develop SBVR from formal documents and later translate it into business process models. In many immature companies, these documents are often unavailable and could hinder resource efficiency efforts. This study introduced a novel approach called informal document to SBVR (ID2SBVR). This approach is used to extract operational rules of SBVR from informal documents. ID2SBVR mines fact type candidates using word patterns or extracting triplets (actor, action, and object) from sentences. A candidate fact type can be a complex, compound, or complex-compound sentence. ID2SBVR extracts fact types from candidate fact types and transforms them into a set of SBVR operational rules. The experimental results show that our approach can be used to generate the operational rules of SBVR from informal documents with an accuracy of 0.91. Moreover, ID2SBVR can also be used to extract fact types with an accuracy of 0.96. The unstructured data is successfully converted into semi-structured data for use in pre-processing. ID2SBVR allows the designer to automatically generate business process models from informal documents. [ABSTRACT FROM AUTHOR]

Semantics of Business Vocabulary and Rules (SBVR) is a standard that is applied in describing business knowledge in the form of controlled natural language. Business process designers develop SBVR from formal documents and later translate it into business process models. In many immature companies, these documents are often unavailable and could hinder resource efficiency efforts. This study introduced a novel approach called informal document to SBVR (ID2SBVR). This approach is used to extract operational rules of SBVR from informal documents. ID2SBVR mines fact type candidates using word patterns or extracting triplets (actor, action, and object) from sentences. A candidate fact type can be a complex, compound, or complex-compound sentence. ID2SBVR extracts fact types from candidate fact types and transforms them into a set of SBVR operational rules. The experimental results show that our approach can be used to generate the operational rules of SBVR from informal documents with an accuracy of 0.91. Moreover, ID2SBVR can also be used to extract fact types with an accuracy of 0.96. The unstructured data is successfully converted into semi-structured data for use in pre-processing. ID2SBVR allows the designer to automatically generate business process models from informal documents.