TextGrid provenance tools for digital humanities ecosystems

摘要

Philological research is very much concerned with the idea of data provenance. However, existing provenance models in eResearch and eScience for data and workflow are ill-applicable to the specific challenges of philological research in the digital age. In philology, provenance data must be collected not only for the complete research objects, but also for fragments — in particular for texts. This necessitates a much richer provenance model that is interwoven with the objects' textual structure and supports the association of provenance data with individual words. It must also permit multiple interpretations of provenance for a single research object. In this, the requirements in digital philology can also be paradigmatic for very fine-granular provenance information collection in other domain-specific digital ecosystems. This paper elaborates key requirements of data and workflow provenance in the philologies and demonstrates how the tools in the TextGrid digital humanities ecosystem answer to those needs. In addition to TextGrid's text-image-link editor and its underlying data model the paper also introduces the humanities' provenance tool HPT for RESTful workflows. HPT is specifically geared to supporting the provenance requirements for workflows in heterogeneous digital humanities ecosystems.