Abstract :
Modern knowledge discovery activities occur in highly dynamic environments. Specific activities may involve multiple tools, techniques, systems, individuals, and locations. In addition to these complexities, the span of time involved with discovery may vary from short to long, as well as being contiguous or disjoint. This paper presents a framework for tracking the history, or provenance, of the discovery process across applications, systems, and users. The resulting capabilities provide fine-grained provenance information relative to the discovered information. Along with the provenance framework, a prototype system is used to demonstrate the main concepts of the proposed approach.
Keywords :
data mining; data visualisation; discovered information; fine-grained provenance information; knowledge discovery; visual exploration activities; visualization systems; Couplings; Data visualization; History; Informatics; Navigation; Organizing; Power system modeling; Prototypes; Space exploration; Web pages;