Output details

Multivalent Digital Object Technology (Fab4 browser)

The Multivalent digital object technology and Fab4 extension is a virtual machine to parse and display digital objects independently of the underlying software and hardware infrastructure. Work at Liverpool on the Multivalent digital object model began in 2005. From 2008 to 2013 the software was extensively redesigned with extensions, and integrated into the iRODS data preservation system.


The engineering challenge was to develop generic parsing technology that could analyse and display meaningful information derived from diverse data objects in a way that is independent of the storage system and that would remain viable far into the future.

The research innovation is a platform-independent virtual machine for interpreting scientific or document data in terms of essential attributes, meaningful structures, relationships, or visual representations. The approach captures the essential attributes of science and document data, and retains it within the archival system and so preserves the ability to interpret and manipulate such digital entities, even within global scale data systems.

The approach allows the viewing and manipulation of documents or other data objects held in any storage infrastructure. This has been demonstrated by using Multivalent / Fab4 software to support different document, media, and CAD formats, serving as the object virtualisation layer. The model also provides augmented annotation capability. This supports shared distributed annotations that are semantically anchored, enabling them to be applied across formats.

The software forms the foundation of multiple international projects, including EU Integrated Projects (SHAMAN, PrestoPrime), the US National Archives (NARA) Transcontinental Persistent Archive Prototype for archiving Congressional records, and the NSF funded DataNet program.