Making Data Meaningful: The Business Intelligence Model and Its Formal Semantics in Description Logics

Business Intelligence (BI) offers great opportunities for strategic analysis of current and future business operations; however, existing BI tools typically provide data-oriented responses to queries, difficult to understand in terms of business objectives and strategies. To make BI data meaningful, we need a conceptual modeling language whose primitive concepts represent business objectives, processes, opportunities and threats. We have previously introduced such a language, the Business Intelligence Model (BIM). In this paper we consolidate and rationalize earlier work on BIM, giving a precise syntax, reducing the number of fundamental concepts by using meta-attributes, and introducing the novel notion of “pursuit”. Significantly, we also provide a formal semantics of BIM using a subset of the OWL Description Logic (DL). Using this semantics as a translation, DL reasoners can be exploited to (1) propagate evidence and goal pursuit in support of “what if?” reasoning, (2) allow extensions to the BIM language, (3) detect inconsistencies in specific BIM models, and (4) automatically classify defined concepts relative to existing concepts, organizing