Making the ANC Digital Archives Work


The ANC Digital Archives, launched on April 29, 2013 at a gala function in Johannesburg, are still a work in progress! In particular, not many of the files that have been digitally captured have had information associated with them such that they can be easily searched for. As a result in hosting the digital archive, Africa Media Online has had to come up with some innovative solutions that allow users to navigate around the archive.
Everything that has been digitised of the ANC Archives is being made available on Africa Media Online’s MEMAT archival digital repository system. Developed by us here in South Africa, we have been working on MEMAT for 12 years and have built it using Open Source programming languages and using open standards. It is now in its third generation and it conforms very closely to the ISO standard for long-term archiving of digital media, the OAIS standard.
 

Search results page on the ANC Archives Research Site
Search results page on the ANC Archives Research Site

 
The real power of the MEMAT system is its management and preservation of media files. Known as the Asset Management Layer, files are placed into Security Boxes in the Digital Vault and once ingested they are check summed and cannot be changed in any way. While files can be superseded, for digital authenticity, the system does not allow a change to the original file. The system also stores all related metadata and can handle any metadata schema that is in use. It can be set up in a distributed manner keeping multiple servers in multiple locations synchronized with one another even in low bandwidth environments.
For the ANC Archives MEMAT is being used to manage all the content generated in the project – photo, video, audio and manuscript files and their associated metadata. It also ensures they are backed up regularly. The system can be made to speak through to any web interface presentation layer including Joomla and Drupal. In the case of the ANC Archives, however,  we are using the MEMAT Presentation Layer that has been upgraded as part of this project to allow for various levels of access as well as time-contrained access. The site has been set up specifically for researchers who want access to the deep archive. A system like this tends to only reach its potential, however, if information has been captured about each digital file.
 
Browse Galleries on the ANC Archives Research Site
Browse Galleries on the ANC Archives Research Site

 
As has been said, the ANC Archives are a work in progress. Not everything in the archive has been digitised. In particular only about a fifth of the manuscripts have been captured in digital form. In addition, of what has been captured, only a small portion has had metadata associated with it. Metadata is information about a digital file. For example, a picture of Oliver Tambo laying a wreath at a grave site to fallen heroes in Mozambique, could not be found on the web site when searching for it unless this information had been captured in a caption and associated with the digital image.
Not having metadata readily available has presented us with some significant challenges in terms of making the archive searchable and accessible. To do this we had to come up with some innovative solutions based on the methodology we used to capture the archive in the first place. The video below introduces you to the system and presents some of the solutions we found for making the ANC Archives accessible in spite of a lack of metadata.