Metaphors To Think Knowledge Graphs By
Knowledge graphs deserve more than just being thought of as IT & DevOps toys. In this piece I explore them as socio-technical phenomena that help us set the stage for new ways of navigating the endless universe of meanings.
These days I am kind of tired to think about knowledge graphs as technical creations only. I am also worried I might lose myself in the data trees, thinking that their sum, dimensions, number of branches and the roots connecting are the wood.
They are not.
Knowledge graphs are only part of a larger ecosystem.
I want to see them as vehicles and further concentrate on the beautiful roads to be traveled with them, not on their nuts and bolts. Or to stay with the woods metaphor, I would rather see them as the “information plumbing” needed to help the Wood Wide Web exchange signals unobstructed.
Just like the Wood Wide Web, BBC showed us in the beautiful video, any content piece on the Web is a signal, bound to connect – to a platform, to a network of data, to a hub of other content.
To open the door for interconnectedness beyond data management and data linking, I thought through several ways of seeing knowledge graphs. These are various artifacts from the noosphere found throughout my work and research in recent years.
And, just an interim note, here I talk about knowledge graphs built with semantic technologies, allowing interoperability, data exchange and the publishing of linked data.
Knowldge Graphs Beyond Data Linking and Data Management, Please
No doubt, knowledge graphs are a lovely, albeit complex, technical tool for solving problems. The need for technical expertise, the right tools and proper project and expectations management throughout their design and implementation are undeniable.
What’s less obvious though is their social aspect. And this aspect is less about technology and more about people. It is less about the formal connections between the data objects and more about the interconnectedness they allow and further foster or untangle.
I hope such a switch of codes will help us non-technical people (what does that even mean?) find our way in the world of knowledge graphs building and usage so that we all can continue to weave together the Web of People, “designed for a social effect – to help people work together – and not as a technical toy” [cit. Sir Tim Berners-Lee, p. 123 Weaving the Web]. And something more, to feel free to “open the way to the critical exploration of an infinite universe of meaning.” (I borrow the expression Pierre Levy’s wonderful post about the digital public sphere).
Knowledge Graphs As A Rhizoma
Origin*: A Rhizomatous Metaphor for Dialogic Theory (Kent, Michael & Lane, Anne (2017) A rhizomatous metaphor for dialogic theory. Public Relations Review, 43(3), pp. 568-578.)
[*Origin in the sense of me, the intertextual animal, reading stuff about dialogue, linked data and other things Cyberia]
Built as a complex assemblage of formalized relationships between data objects – events, people, books, locations, web pages, products, to name a few, knowledge graphs allow interaction with the knowledge codified in their system.
And it is that interaction with priorly formalized knowledge, that branches out constantly, fostered by the discovery of hidden relationships or by the emergence of novel ways of relating previously unrelated things. This depth of the exploration of connections invites a parallel with the rhizomatous metaphor.
In an article, called A Rhizomatous Metaphor for Dialogic Theory, Kent and Lane present an alternative metaphor of dialogic communication, contrasting a rhizomatous metaphor to the arboreal analogy often used in communication theory. Under the arboreal metaphor, humans make choices among competing alternatives, rather than collaborating on contingent truths and solutions. Choices are bound by rationality rather than cocreationally constructed, write Kent and Lane. In contrast, the rhizome, allows a perspective on communication, more specifically on dialogic communication, where what is happening on (or in) the ground is the real action.
Rhizomatous exchanges occur seemingly randomly and haphazardly as participants actively seek contact from others who might not be identified by more linear (arboreal) lines of thought or ways of researching. This almost anarchic approach to identifying interlocutors, and circumventing traditional communication channels that inhibit free and open communication, are vital points of difference between the metaphors of the rhizome and the tree.
cit. Michael L. Kent, Anne B. Lane, A rhizomatous metaphor for dialogic theory.
The rhizome researchers used is also based on a chapter called “Rhizome, by Gilles Deleuze and Félix Guattari where they introduce the rhizome as a metaphor for knowledge or understanding, describing how, like a rhizome, culture and knowledge spread through ceaselessly established connections.
To drive the point home, Kent and Lane also provide their view of the monologic and dialogic way of communicating with the public.
The reason I find that so apt when we talk about knowledge graphs from the perspective of how we can utilize them for communication is the need for dialogue that emerges when building them. And all the conversations that branch out when thinking about definitions, connections, hierarchical relationships, ways of naming things, and other taxonomies and ontologies adventures.
All of these questions are now just a paragraph running on your screen, but when you put that into action among many stakeholders with different backgrounds, there come the rhizomatous nature of knowledge. People are different, their background and mindset are different, they way they see the world is different. And this is a variety not to be unified but rather incorporated in a knowledge graph. Last but not least, for sure there the arboreal metaphor would be more apt for enterprise knowledge graphs, and the rhizomatous metaphor would rather suit the folksonomy nature of Wikidata.
Knowledge Graphs Are Curiosity Cabinets
Origin*:The Anotomy of Curiosity
[*I first stumbled upon curiosity cabinets in literary theorist, essayist and translator Michał Paweł Markowski’s book The Anatomy of Curiosity. ]
I learnt about curiosity cabinets from the book “The Anatomy of Curiosity” where author Michał Paweł Markowski offered a thought-provoking journey throughout the human desire for knowledge and understanding. Further I explored knowledge graphs as curiosity cabinets in this article for Ontotext from the perspective of the digitization of the world’s cultural heritage (ref. If Curiosity Cabinets Were Knowledge Graphs)
In essence, curiosity cabinets were “small collections of extraordinary objects which, like today’s museums, attempted to categorise and tell stories about the wonders and oddities of the natural world”. https://www.bl.uk/learning/timeline/item107648.html
I see curiosity cabinets, them as a useful metaphor of the intellectual exploration not limited to a single area of interest, but rather celebrating the joy of discovery and the interconnectedness of things and the ability to recontextualize artifacts (and thoughts). Just like knowldge graphs.
A beautiful way to illustrate these semiotic routes, enabled by data, I am talking about, is this video by Research space.
In their aspect of serving as a repository of interlinked objects, not only between themselves but also with outer world knowledge, e.g. connected to other public knowledge graphs like Wikipedia, knowledge graphs open the door for new semiotic interpretative routes and novel ways of seeing objects, just like curiosity cabinets did. Only that with knowledge graphs, the ones built with semantic technologies, you can follow your nose through a limitless cabinet: the Web.
A Knowledge Graph is A Molecular Model, Just like Text…
With my philology background and love for intertextuality I can’t help but see knowledge graphs as giant texts reaching out and weaving into each other. Texts with rich metadata describing them and the concepts they refer to and play with are for me an emerging form of textuality – one rooted in wider access to information and greater opportunities to exchange thoughts through shared vocabularies.
Sarven Capadisli’s project dokieli is just that. Exploring the principles and the interface of Dokieli can help us understand how Linked Data (and I see knowledge graphs as assemblages of Linked data) enriches the communicative power of a text, but also its heuristic powers.
When I first saw the project, which is now evolving as Solid project, is, it brought me back to what Negroponte wrote about text in Being Digital:
An expression of an idea or train of thought can include a multidimensional network of pointers to further elaborations or arguments, which can be invoked or ignored. The structure of the text should be imagined like a complex molecular model. Chunks of information can be reordered, sentences expanded, and words given definitions on the spot. These linkages can be embeded either by the author at “publishing” time or later by readers over time.
Being Digital, N. Negroponte.
In the light of understanding a knowledge graph, and thinking though it’s use for, “content” – sorry for the word, let’s not forget that “content” is nothing but an artifact of thinking, discussions, collisions, inner relationships, traces of processes of reaching shared understanding, this allows us to understand how we can use a knowledge graph to underpin immersive experiences with text. And underpinning these experiences we also build the Giant Global Graph.
The way I am connected, not the way my Web pages are connected. We can use the word Graph, now, to distinguish from Web. I called this graph the Semantic Web, but maybe it should have been Giant Global Graph! Any worse than WWWW? 😉
Giant Global Graph, Tim Berners-Lee
The Assemblage Theory and Knowledge Graphs as Rituals of Relating
The epilogue of this blog post is actually the beginning of a hypothesis. This came when I was working to communicate the idea about data and dialogic marketing communication.
It seems to me that a good way into understanding (and building better and richer) knowledge graph is perceiving it as an assemblage of Linked Data.
The assemblages theory has it that:
An assemblage is comprised of objects and their connections, which combine to make up interconnected arrangements with their own functional properties and capacities. An object can be anything that has an effect on the world: humans, technology, animals, policies, or opinions. An assemblage can be any arrangement of objects: a football team, a zoo, a large-multinational, or a language classroom. Key to an assemblage is its co-functioning; that an object’s capacities only become realised in relation to other objects. For example, a teacher may use a mobile device to revise a particular language point. The technology on its own does not have the capacity to be used for language teaching.
cit. Assemblage theory: coping with complexity in technology enhanced language learning
To extrapolate that into the place of knowledge graphs, their capacities only become realized in relation to other objects, to other people. And this is why we need more people of different backgrounds to walk the talk about knowledge graphs and enact new rituals relating to content, data and collaboration.
Knowledge Graphs Need Brave Interdiciplinary Teams
Having a rhizoma, a molecular model and a curiosity cabinet in your organization takes brave interdisciplinary teams willing to change practices, enrich workflows to encompass data depths of content, and most importantly transform ways of seeing knowledge. From my sand tower, building knowldge graphs is related to many purely dialogic experiences rooted in the understanding of the heteroglossia present in organizations and the need for co-orientation (concepts I explore in my upcoming book Being Dialogic).
And this is why we often need to step back from the technological automagics of knowledge graphs into their essence of being vehicles to travel and discover unknown roads by.