Visual analytics is a costly endeavor in which analysts must coordinate the execution of incompatible visualization tools to derive coherent presentations from complex information. Distributed environments such as the Web pose additional costs since analysts must also establish logical connections among shared results, decode unfamiliar data formats, and engage with broader sets of tools that support the heterogeneity of different information sources. These ancillary activities are often limiting factors to our vision of seamless analytics, which we define as the low-cost generation and reuse of analytical resources. In this paper, we offer a theory of analytics that formally explains how analysts can employ Linked Data to maintain and leverage explicit connections across shared results as well as manage different representations of information required by visualization tools. Our theory builds on the well-known benefits of interconnected data and provides new metrics that quantify the utility of interconnected user- and task-centric, analytical applications.
To describe our theory, we first introduce an extension of the W3C PROV Ontology to model analytic applications regardless of the type of data, tool, or objective involved. Next, we exercise the ontology to model a series of applications performed in a hypothetical but realistic and fully-implemented scenario. We then introduce a measure of seamlessness for any chain of applications described in our Application Ontology. Finally, we extend the ontology to distinguish five types of applications based on the structure of data involved and the behavior of the tools used. Together, our seamlessness measure and application ontology compose our Five-Star application theory that embodies tenets of Linked Data in a form that emits falsifiable predictions and which can be revised to better reflect and thus reduce the costs embedded within analytical environments.