Analysis of approaches for supporting the Open Provenance Model: A case study of the Trident workflow workbench

The Trident workbench is a platform for composing, executing and managing scientific workflows. While Trident collects provenance in its native provenance model, the third provenance challenge was an opportunity to build support for the Open Provenance Model into Trident. There are several possible approaches to harmonize our native model with OPM, and such choices are also available to other existing provenance and workflow systems working towards OPM compatibility. We identify and analyze the relative merits of these approaches in an effort to inform practitioners planning to support OPM in their existing provenance/workflow systems. Further, we describe our experience with using the integration approach we choose to interoperate with other teams as part of the challenge.

Research highlights
► Practical approaches to support OPM in existing provenance systems are less studied.
► Tight coupling of OPM into existing system gives uniformity but has coding overhead.
► Storage coupling maintains both native and OPM models but has storage overhead.
► Query coupling builds OPM views on native model for low overhead but need for DB.
► Trident decouples native and OPM models to meet the third provenance challenge.