Over the last couple of weeks a really interesting discussion has been taking place on the Early Modern Online Bibliography blog run by Eleanor Shevlin (West Chester University of Pennsylvania) and Anna Battigelli (SUNY Plattsburgh). Based in the US, Eleanor heard about JISC Historic Books and its use of meaning-based searching rather than traditional keyword searching enabled by the use of Autonomy IDOL software and wondered:
“Over the past three or four years (and maybe longer) I have been consistently struck by the transformations that traditional searches of ECCO, Burney, EEBO, as well as Google Books have had on the ways I think about searching, construct searches, and view my results. More specifically, these keyword searches, described here as traditional, were already encouraging me to view results in a more networked, contextual way and, as a consequence, to devise additional searches aimed at teasing out new potential relationships. The meaning-based search enabled by JISC’s mimas platform, of course, is offering something quite different, but I wonder how its use might cause rethinking of what it means to search and research.”
Quick to help Eleanor understand more about how JISC Historic Books works was Dave Mazella and Kelly Centrelli on The Long Eighteenth blog. Kelly posted a really useful overview with images of the ‘conceptual search’ enabled on JISC Historic Books and how the results differ to ECCO on the Cengage platform.
Following Kelly’s post, a further discussion ensued on the EMOB blog around the differences between ‘search’ as offered on the Cengage platform ‘for retrieval of known terms’ and how, as Kelly suggested in her post, JISC Historic Books ‘enables search and discovery’. The difference of the two different search methodologies was summed up by Kelly:
If I were to compare the two ECCOs to things an advisor may say, I imagine they would sound like this: A JISC search tells me, ‘Good job! But did you think of this, and this, and this?’ A Cengage search (at least in my experience) tells me ‘Nope. Think harder, and try again.’ A bit anecdotal, perhaps, but depicts how the different approaches not only work, but how they may resonate with a user. Moreover, it alerts us to a key problem; that is, how much do we really know, think, and process, when we’re always clued in to the answers? And, if we think we are clued in to these answers, will we ever try to think beyond them? Beyond the suggestions given?
What Eleanor and Kelly are discussing in their posts is what impact these approaches have on research and the searcher’s behaviour. Certainly something that could be explored in more detail? I’m going to bring it to the attention of the JISC Historic Books advisory board and see what they have to say. In addition, we will be working with Mimas to help user’s understanding the Autonomy software and how this works on the platform.
In the meantime, I would like to thank Eleanor, Kelly and colleagues for their blog posts and look forward to seeing more discussions in this area.
Here are all the links you need to follow the conversations:
Other posts of interest: