A Semantic Web Management Model for Integrative Biomedical Informatics

xlxzj

xzlzx

It has not been lost to the swelling ranks of Systems Biologists that the reduction of data interoperability to the ternary representation of relations [34] brings the topic solidly back to the Galilean fold of Mathematics as a language. The reduction of data structures to globally referenced dyadic relations (functions of two variables), such as those of the Entity-Relationship (ER) model, brings in rich feeds from the vein of Logic. In the process, and beyond Galileo's horizon, assigning a description logic value[35]–[37] to some RDF predicates (for example, specifying that something is part of or, on the contrary, is distinct from something else) allows the definition of procedures. This further elaboration of RDF has the potential to transform data management into an application of knowledge engineering, and more specifically of artificial intelligence (AI). This reclassification reflects the dilution of the distinction between data management and data analysis that is apparent even in an introduction as brief as this one. Another clear indication of this transformation is that it re-ignites the opposition between data-driven and rule-driven designs for semantic web representation[38]–[42], a recurring topic in AI. It is important to note that the management model proposed here is orthogonal to that discussion. Its purpose is solely to enable the distribution[43] of a semantic data management system that can withstand changes in the domain of discourse, independently of the rationale for the changes themselves.