Antwerp, December 2021, Carlos Tubbax.
The value proposition of EKG¶
Companies face increasingly difficult challenges due to the complexity of their business environments stemming from market disruptions, new regulations, the COVID-19 pandemic or other large scale events. In such a challenging business world, being able to have a clear oversight of an organization as a whole has become more critical than ever to assess risk exposure, to find supply chain bottlenecks, to comply to regulations among other business requirements. Therefore, connecting the dots spread inside and outside each company is necessary for all these crucial tasks. In this context, the word dots refers to data needed to carry out these different tasks.
However, connecting the dots is far easier said than done as these dots are most often spread across a myriad of business silos that may not even aware of each other’s existence and in different formats that make the task of integrating and harmonizing data across the enterprise overly complex.
As a solution to this problem, the Semantic Web and its underlying standards were developed as an answer to integrate, identify, catalogue, harmonize and assign meaning to individual pieces of data from different sources. Enterprise Knowledge Graphs (EKGs) aim at doing that at an enterprise-wide level by using Semantic Web standards and technologies. Additionally, EKGs can boost other technologies such as machine learning and blockchain by feeding them with high-quality and meaningful data.
The purpose of this maturity model is to guide companies and other organizations throughout the development of Enterprise Knowledge Graphs as a tool to develop capabilities that they may consider vital for their own businesses such as risk management or customer 360° and for which connecting the dots is necessary. Since this journey will certainly require lots of time and effort, this maturity model aims at providing a step-by-step rationale of how companies can develop EKGs that serve their own business needs. The maturity model will guide enterprises throughout its different maturity levels that will mark the progress of the Enterprise in developing EKG(s) for this kind of enterprise-wide use cases. In order to make the utility of Knowledge Graphs for enterprises clearer, this part will provide some examples of how knowledge graphs could be used to tackle the aforementioned problems and to connect the dots for some use cases. However, the decision of which use cases need to be supported and prioritized is up to each enterprise. Therefore, the following examples are just mere illustrations of what companies can do by connecting the dots.
People using Amazon’s Alexa, ordering food through Uber Eats or booking a listing on Airbnb are using a knowledge graph even if they are totally unaware of that. Amazon, Alexa, Uber among other cutting-edge companies have been using knowledge graph technology to disrupt entire sectors and to create new value propositions to their customers.1 For example, Amazon uses a product graph to categorize products on its retail website and to make better product suggestions to customers among other things.2 In that context, Knowledge Graphs fulfills the role of disruption enabler. In order to illustrate this, it will be discussed how Airbnb uses a knowledge graph to enhance its customer experience by integrating and conducting computer reasoning on data coming from different data sources.
On the other hand, Enterprise Knowledge Graphs may also provide great value when being used to support the daily operations of a business and, in that context, the enterprise is their end user. Although there are fewer examples of knowledge graphs being used in this context, Deloitte’s white papers discusses how knowledge graphs could be used in organizations to unravel the intricacies of their own business processes, relationships, supply chains, etc. by the means provided by Enterprise Knowledge Graphs as a business supporting/enabling technology.
Customer experience enhancement at Airbnb¶
In order to move towards its vision of becoming an end-to-end travel platform, Airbnb needs to be able to provide customers insights that help them decide when to travel, where to travel and what to do in their trips. For instance, Airbnb needs to be able to answer queries such as:
- What are the most popular landmarks and neighborhoods in London?
- Which Airbnb listings are best suited for working nomads?
- What are the most popular Italian restaurants in New York?
Answering these queries may help travelers plan their trips much better and, in turn, Airbnb may increase its customer value proposition.
As a means to answer all these customer queries, Airbnb uses a knowledge graph represented by the figure on the right.
As explained by Chang (2018)3, semantic web knowledge graphs
offer the ability of structuring and adding meaning to data
from different sources such as relational databases in a scalable
way to answer these queries.
Additionally, it is far simpler to connect data about a
certain object spread across different data sources with a
knowledge graph than it is with a relational database.
Another advantage of knowledge graphs compared to
relational databases is that new relationships between concepts
or objects can be added to a knowledge graph in a much more
flexible and simpler fashion.
For instance, knowledge graphs can link the concept surfing as
a sport to the concept surfing which Hawaii is known for.
So, if a customer asks Airbnb’s website for surfing destinations,
Hawaii may appear on the query results.
This kind of associations might be far more difficult in a
relational environment especially if the data about them
come from different sources with different schemas.
By connecting data about places, experiences, homes, restaurants, etc. as illustrated here, Airbnb is able to answer complicated user queries4 such as What events, restaurants and homes are located in Mission District in San Francisco? as shown on the right and below.
In this way, Airbnb is moving forward into becoming an end-to-end travel platform that serves its customers throughout their entire journeys instead of only renting listings out.