A long time ago, in a place quite near, I worked with Fleur and Rob on making a server based ‘Hyperwords’ system for the Chinese NBA (National Basketball Association) website through Solina, which I have put up a brief video capture of: youtu.be/kZmhbmv8Yfs.
We built the system to show a graphic when the user pointed to a player or a team which should instantly and visually answer the question of whether the player or the team is any good–how they are doing now and how they have done historically. This was not a difficult design task but of course it took much work and iteration to get it right. The problem was with the data API and collaboration with its provider.
For my PhD I am doing something very similar: When a reader comes across a citation in a document or in the References section, how can the system answer the question: Is this useful? And if it is, how can I most effectively access the document in a way where I know where I found it and see its further context?
Human recorded information is prone to error and bias so it turns out that the simple act of specifying the identity of a citable document, whether book or academic article can turn out to be messy and therefore this needs to be designed into the system. As many have suggested before (Nelson, Gutteridge etc.) a way to easily tunnel in and out of documents and contexts can be useful.
All information is liquid but our tools are not.
When coming across a citation in a document a mouse-over which performs searches to see further information about the article itself, such as the abstract and the authors, could quickly and effectively help the user decide whether to download the article.
I’ll therefore be looking at the ACM Digital Library dl.acm.org which has a well structured corpus of information about documents and authors. This should be somehow scrape-able for what we are talking about, for a proof of concept. For example, someone cites ShyWiki: a spatial hypertext wiki prototype (Solis & Ali, 2008) and the user can click on this citation for an expanded citation dialog:
The user should be able to click on the name of the document to open it on the site to download it and more.
It’s a simple thing, but could it be useful? Could further interactions come from this approach?