ContextMiner: Supporting the Mining of Contextual Information for Ephemeral Digital Video Preservation

Emerging information media present new challenges to the curators. While archiving objects, and building meaningful collection for long-term preservation and access, have been well-understood practice for centuaries, digital objects present new issues. In the previous article (Shah, 2009) I identified a number of these issues related to digital objects, specifically digital videos of an ephemeral nature. I argued that while preserving such objects, adding contextual information is essential. One of the interesting challenges is to identify what to collect and preserve as contextual information. For ephemeral digital videos, I proposed to harvest four kinds of relevances and five kinds of contexts. In order to implement this proposal, I presented ContextMiner, a framework and a system to support digital video curation. In this article, I will take a closer look at ContextMiner, analyzing it for its functionalities and usability. This is done by usability inspection and content analysis. For the former, we simulated two curatorial tasks, asked our users (curators) to use ContextMiner, and provide us feedback on its usability and functionalities. For the latter, we mined a collection prepared by ContextMiner for its potential usage in preservation. Finally, I have summarized the lessons learned from developing and using our system, providing implications for digital library curators interested in collecting and preserving digital objects of an ephemeral nature.