1 Wait, What The Heck Is A Knowledge Graph?

I can give you a bunch of academia and business definitions of knowledge graphs with detailed explanations and use cases and you will still be puzzled and without a clue about what exactly is a knowledge graph and how it works. A very recent explanation I found in the wonderful paper Generations of Knowledge Graphs: The Crazy Ideas and the Business Impact by Luna Dong^[1] – Principal Scientist at Meta Reality Labs, leading the ML efforts in building an intelligent personal assistant^[2], reads:

“Knowledge Graphs (KGs) have been used to support a wide range of applications, from web search to personal assistant […] KGs model the real world in a graph representation, where nodes represent real-world entities or atomic (attribute) values, and edges represent relations between the entities or attributes between entities and atomic values. A piece of knowledge can be considered as a triple in the form of (subject, predicate, object), such as (Seattle, located_at, USA). The data instances in a KG follow the ontology as the schema, which in itself is represented in a graph form and can be taken as a part of the KG. The ontology describes entity classes, often organized in a hierarchical structure and also called taxonomy, and meaningful relationships between classes.”

As you can see, knowledge graphs are really complex technology with lots of moving parts and new concepts to grasp, given you, like me, are a non-technical person, without data modeling, data harmonization and entity linking knowledge and practical experience. And although there has been a great uptake in marketing content (me being part of that effort and guilty as charged^[3])  that explains and promotes knowledge graphs to wider audiences, knowledge graphs do remain something that takes time and background knowledge to fully grasp.

As a starter, I chose to share with you a very accessible illustration by prof. Elena Simperl of King’s College London. Asked to explain knowledge graphs and the nodes and edges they are built of, prof. Simperl grabbed a piece of greenbar computer paper and drew a node to denote Bush House, further drew an edge and then connected the Bush House to a picture from a brochure lying on the table^[4].

And this is the simplicity at the end of complexity when it comes to linking data from heterogenous sources. Knowledge graphs are this – a structure to represent connected entities from different sources, expandable and easy to link one to another. Just like prof. Simperl did. In the video you will see how she builds on top of the graph, taking from what’s around – it’s like she puts her cognitive semantic network to paper, literally, to a graph. Drawing a line to an image from a newspaper, she is then naming the connections and expanding the representation. In technical terms that might well translate into having one schema – ontology which defines the entities you want to describe in your graph and then taking from the outside world, let’s say from Wikimedia, some images that are related to your entities.

All project-management hurdles, company-wide efforts of synchronization, and technological conundrums in the building of a knowledge graph (and I won’t lie, there are many) aside, knowledge graphs are collections of data representing interrelated facts.

Typically, knowledge graphs include networks of computer-readable, interlinked representations of entities, such as people, places, events, concepts, things. As complex systems based on structured data and enabling the creation and management of a wide range of technological solutions by making them more intelligent, knowledge graphs are a programmatic way of modeling knowledge domains using subject matter experts, interconnected data, and machine learning algorithms  A knowledge graph can be built either with or without the use of semantic technologies by integrating data^[5].

As we saw, the term knowledge graph gained popularity after 2012, when Google introduced their Knowledge Graph. Researchers Lisa Ehrlinger and Wolfram Wöß refer  to the term as a buzzword adopted by business and academia alike to describe different knowledge representation applications^[6]. Knowledge graphs  are used by software companies such as Google, Microsoft and Amazon in the development of smart technologies, in the creation of dialog systems, personal assistants and artificial intelligence. Being interconnected repositories of data, they are also utilized by large corporations such as IBM, Samsung, Ebay, Bloomberg, NY Times, Twitter^[7]. Among the solutions that knowledge graphs provide are enterprise knowledge management, data management in healthcare, management and description of digital resources in the field of cultural heritage. In 2019, the European Union launched the “KnowGraphs” project^[8], whose goal is to make this type of architecture of connected data more accessible to a wide audience of companies of different sizes and professionals in different domains.

Among the known enterprise knowledge graphs are those of Bing, Google, Airbnb, Amazon, eBay, Uber, LinkedIn, Accenture, Bloomberg, Capital One. For a detailed review, enjoy this lovely book available online^[9].

There is also a larger body of publicly available knowledge graphs, such as DBpedia or Wikidata. You can compare knowledge graphs on the Web (think repositories that your own knowledge graph can link to, thus enriching its contents) in Knowledge Graphs on the Web – an Overview^[10], where authors list and compare publicly available knowledge graphs, giving insights into their contents, size, coverage, and overlap^[11]. Open knowledge graphs are published under the Open Data philosophy, where open means anyone can freely access, use, modify, and share for any purpose. Open knowledge graphs include DBpedia, Freebase, Wikidata, YAGO.

Along with the concepts of open and enterprise knowledge graphs, there is an emerging concept looking to point to the representation of structured information about entities that are personally important to a given user and with which they interact on a daily basis. Such entities are traditionally siloed in documents, notes, app data etc tools we use everyday. Personal knowledge graphs^[12] are also very much like personal  data spaces and are conceptually close to what Tim Berners-Lee envisions in his project Solid^[13]. We will not talk about it in detail, yet I encourage you to keep an eye on its development, as this is something that we as marketers will one day meet, in one way or another. Think a world where everyone is in control of their data, and shares it only with selected few companies, whom they trust.

A Non-technical Perspective Towards Knowledge Graphs

Being a philologist and curious about the language existence of organizations, I do read a lot of technical documents about knowledge graphs and their development. That’s because I see these technological solutions as repositories capable of enabling better communication of concepts and messaging both within the company and outside it. Just because they allow the use of consistent vocabularies and the creation of common data layers.

Yet, if I am to stay with the larger question, beyond the technicalities of knowledge graphs, into the “So what for us, as intertextual beings?” narrative, I would describe knowledge graphs as curiosity cabinets.

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. 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”^[14].

I see curiosity cabinets as a useful metaphor of an 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^[15]).

Just like knowledge graphs I have explored knowledge graphs as curiosity cabinets in this article^[16] from the perspective of the digitization of the world’s cultural heritage.

Knowledge graph technology can walk us out of the lack of context (which is basically absence of proper interlinking) and towards enriching digital representation of collections with semantic data, which is further interlinked into a meaningful constellation of items.

With knowledge graphs, additional facts and figures can be threaded into the collection items and the metadata related to them. Imagine a curiosity cabinet with items attached to threads (strings) of well-described semantic information, linking them to other artifacts, events, people, institutions, you name it. Then, we would have at our fingertips a collection rich in connections that can be further explored, depending on the interest of the viewer.

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 with knowledge graphs, the ones built with semantic technologies, you can follow your nose^[17]  through a limitless cabinet: the Web.

Great, But So What For Marketing Communications?

Following their nose, the people we want to connect with through our marketing communications, are traveling a myriad of cyber spaces. In these information-intensive environments on the Web, our marketing communications compete with all kinds of content – content generated by users, content often from the communication strategies of other organizations, content resulting from individual and organizational communication.

Given such “travels” saturated with messages, the paradigms of marketing communications, their strategies and structures are evolving so that knowledge becomes a competitive advantage^[18].

Knowledge graphs could function as such a system, similar to what Rashi Glaser theoretically described as a knowledge base. They have the potential to be a means of creating a continuum of interactions in an information-intensive environment. Such architectures are highly needed as our marketing communications on the Web are highly dependent on knowledge-based interaction – knowledge of products, services, market environment, consumer context. With myriads of datasets available on the Web and hundreds of platforms and ways of connecting, the accumulation, storage and easy access to this knowledge is impossible without the help of a system that can serve as a system for managing the artifacts of this knowledge, namely the marketing content, user interaction history and data related to both.

Seen from the perspective of creating marketing communications, both open and enterprise knowledge graphs can serve as a living system helping marketing communications professionals do their knowledge-intensive job better. Knowledge graphs can also assist people in looking for solutions on the Web, providing interconnected data and … content.