I don’t have access to the EBSCO Discovery Service, but above is an example screenshot provided by EBSCO as an example. A concept map (or knowledge graph) is just a visual collection of links. Sometimes, as here, there is no differentiation between the relative importance of the links. Other concept maps show the relative number of resources available via the link, for example, the OECD SDG Pathfinder (which uses UNSILO concept extraction). The diameter of the circles corresponds roughly to the number of articles tagged with each concept:
Concept maps look great. As Bresman states, “since the maps essentially show users knowledge graphs, they make it easier for users to make connections across topics”. So why are they not more used? And are they really, as Bresman claims, “trekking into the semantic frontier”? I would hesitate to call them semantic. He states “users can find hidden relationships between and among concepts”. What he means is that a tool behind the scenes has found some links between articles or chapters, and is now displaying that link. I can’t deny that some of the links discovered are fascinating (see the screenshot above).
Above is an example of Yewno Discovery. I see the idea: start from a concept such as “social media”, and think of all the related topics. But the topics shown are neither comprehensive nor very meaningful: some are obviously incomplete. There is a link to Twitter, but no link to Facebook, or LinkedIn. There is a link to “online tutoring” which I would think has only a marginal relevance to social media.
Perhaps a paradox of the present day, when the world seems to be moving away from text towards voice and visual cues (think Alexa and Audible rather than Google and e-books) is that some visual-based presentation systems simply don’t provide a meaningful way of understanding the content. Television documentaries, for example, are weakest when they provide statistical or graphical material.
I suspect that many concept maps are examples of technology-led solutions. Because the technology can provide such visualisation tools, the supposition is that we should all adopt them. Perhaps we become suspicious of anything more detailed than a tag cloud – and, after all, the Wikipedia definition of a tag cloud is “a novelty visual representation of text data”. “Novelty”! But I think they are right. Nobody has come up with a serious use of it to date. We are still searching for a genuinely useful visual representation of relationships for text content.