Taming Academic Text: The Open Research Knowledge Graph
Taming text is hard, taming scientific knowldge is even harder. But what proved the hardest was taming my enthusiasm while preparing this sui generis report about The Open Research Knowledge Graph (ORKG). Yet another graph I have the pleasure of exploring and presenting to you, of course, always looking at the endeavour with my web content and dialogic communication lover’s inner eye.
Not only the ORKG opens so many possibilities and is such a thoughtful, methodological approach of building a graph of scientific knowledge, but also it is a project that can guide us all towards the ways we build knowldge graphs with people and processes in mind, embedded in the very fabrics of the platform and its technological wrarps and wefts.
I won’t hide, I am fascinated and I wish I could have something readily-available for content – to show my students and people who ask me – okay, how to write for the Web . Don’t write for the Web, I would tell them, write to unlock knowledge and then all else will fall into place, graph, I mean (the type of https://en.wikipedia.org/wiki/Giant_Global_Graph).
That passion of mine shared now, let’s take a step back out of my dreams about meaningful marketing content and into the real world of the the beautiful Open Research Knowledge Graph and its mission to move scientific knowledge forward – from PDFs towards FAIR data (data which meets the FAIR principles of findability, accessibility, interoperability, and reusability).
A Web Of Knowledge. Linked.
Among the many vivid examples of how the Semantic Web is a sea change to the ways we work and live, there was an example of a yellow taxi, I remember Sir Tm Berners-Lee refering to in the book Weaving the Web. Searching through my dog-eared copy of Weaving the Web, I couldn’t dig it out, though. What I could however happened to be even better an illustration of the story I want to tell you in this edition of Knowledge Graphs ❤️ Content (Stories about knowledge graphs powering content experiences). It is the following quote:
People would also have a running model of their plans and reasoning. A web of knowledge linked through hypertext would contain a snapshot of their shared understanding.
p. 162. Weaving the Web: The Original Design and Ultimate Destiny of the World Wide Web by its inventor. Tim Berners-Lee
What a great way to explain what The Open Research Knowledge Graph (ORKG) does.
A Knowledge Graph of FAIR Academic Research To Talk To
I have been following ORKG since its first public announcement. It was back in 2019 when Dr. Sören Auer (Director of TIB) gave a talk titled: Towards Knowledge Graph-based Representation, Augmentation and Exploration of Scholarly Communications.
In short, the graph aim is to describe research papers in a structured manner, making academic papers easier to find and compare. Not only that but the project also launched ORKG Ask (https://ask.orkg.org/) where we can browse scientific research, via scholarly search and exploration system powered by Vector Search, Large Language Models and Knowledge Graphs.
The ORKG makes scientific knowledge human- and machine-actionable and thus enables completely new ways of machine assistance. This will help researchers find relevant contributions to their field and create state-of-the-art comparisons and reviews. With the ORKG, scientists can explore knowledge in entirely new ways and share results even across different disciplines.
https://orkg.org/about/1/Overview
The Machine-Actionable Way Forward
The Open Research Knowledge Graph (ORKG) helps solve the problem of text-heavy, unstructured, quasi-digitized research documents by offering an online platform where scientific knowledge is shared in a structured, machine-readable way. It supports researchers in creating, organizing, and using information that’s easy to search, compare, and build on, saving time and improving collaboration.
To build ORGK its team use a “a mixture of manual (crowd/expert sourcing) and (semi-)automated techniques.” (cit. https://www.degruyterbrill.com/document/doi/10.1515/bfp-2020-2042/html ) It is with this synerge that they aim to achieve quality and reliability of the graph.
The technologies put to work for ORKG allow scholarly knowledge to be view from different angles and sliced and diced addressing particular research quests.
Among the main features of the knowledge graph are interactive charts, comparisons, natural-language Q&A and other emerging features that guide scientific research journey, lit by the Semantic Web vision.
Very long more-than-5-year-old research short, ORKG allows the exploration of scientific knowledge through:
- Comparisons, where research ontributions are displayed side-by-side, letting the user to compare results, methods, and more at a glance.
- Reviews which are community-maintained surveys that stay up-to-date and are machine-readable, allowing the user to access “live” knowledge.
- Lists – a curated collections of relevant literature allowing users to easily collect and organize papers around a topic to support collaboration.
- Templates – a stort of structured authoring type data entry and support automation for research insights, allowing user to standardize how content is added and help train AI to extract information automatically.
- Structured Descriptions through which papers become semantically rich, machine-readable data where key elements of a scientific research are broken down in components: problems, methods, and results.
- Observatories (what a cool framing! And a nice human-in-the-loop element) – ORKG Observatories organize research contributions in a particular research field and are curated by research organizations active in the respective field. There’s plenty of them now: https://orkg.org/observatories
Read more at: “Towards Knowledge Graph based Representation, Augmentation and Exploration of Scholarly Communications” where Sören Auer provides an overview of the ORKG initiative. But more importantly get involved: https://orkg.org/about/21/Get_involved
Also, you can follow your nose and dive into the features through the guide ORKG team create here: https://orkg.org/about/14/Get_started.
Here, I compiled a list of resources to dive into:
- ORKG Success Stories
- Improving Access to Scientific Literature with Knowledge Graphs
- Increasing Reproducibility in Science by Interlinking Semantic Artifact Descriptions in a Knowledge Graph | SpringerLink
- From Papers to Knowledge: Representing Scholarly Contributions in the Open Research Knowledge Graph (ORKG) – TIB AV-Portal
What if Website Data was FAIR: Tim in FAIRland
Imagine the above features applied to an enterprise content repository, open to browse by the public.
Wouldn’t that be the dream of a content writer and any content person – to be able to give their users – a web of deeply interconnected information about a given topic. And a web that is not the thing, the interaction of the user with it is.
Just like researchers who “who must sift through vast amounts of information to derive meaningful insights” we, the users of the Web and its riches, need to dig through giant cyber fields of information when looking to learn more about a given product or service.
To stick to my imagined user Tim and the benefits of a content knowledge graph, we can imagine a platform like ORGK being truly helpful for an enterprise to present their knowledge to a stakeholder.
I would imagine Tim, no matter whether he is using the platform as a internal stakeholder or a prospect, to navigate knowledge graph objects rather than monolithic webpages.
But how would that look like?
In synch with the concept of allowing the user to pull information from the website, navigating a platform like ORGK would allow Tim to:
- Talk to the website content
- Compare pieces of information (of course that would involve not only a knowledge graph technology but also a way of structured authoring)
- Enjoy, as an ORKG paper frames it “Machine-actionable” information.
The latter is paramount to the upcoming years of marketing. Take Visa’s Enabling AI agents to buy securely and seamlessly https://corporate.visa.com/en/products/intelligent-commerce.html.
If we truly imagine “a future where an AI agent can shop and buy for you” then we would also need to reimagine the way we create and publish content on the Web.
And there’s a ton to learn from ORKG and their larger scope of redefining the way scientific knowledge is represented, published and used. shape a future scholarly publishing and communication where the contents of scholarly articles are FAIR research data.
This knowledge graph story is part of the stories I told in Being Dialogic – my book about marketing communications on the Web and the way we weave enterprise content online, through dialogue and metadata.