Presented by Stardog
Chances are, you’ve already interacted with a knowledge graph today. If you’ve searched on Google, asked Alexa a question, searched for someone on LinkedIn or Facebook, or ordered food on UberEats, you have! What all these products have in common is they are all powered by a knowledge graph, a data management tool that captures and organizes the relationships and associations between concepts, allowing companies to build more human-centered products and manage all data as a connected whole.
Knowledge graphs make data not only machine-readable, but also machine-understandable, by capturing real-world context that is so often MIA. After all, everyone has had a bad experience with Siri because sometimes, machines just don’t get it.
Let’s look at Uber Eats as an example — the delivery service was able to capture user intent by building a knowledge graph into their product. They saw a difference between search queries and orders but recognized that these differences were often the result of the subtle nuances of how human brains connect concepts. For instance, what happens when a hungry user searches for a particular cuisine when dinner time rolls around? An eater might search for one dish but end up ordering another related dish. For humans, the similarities between different dishes is obvious, but for machines, this is not necessarily the case, especially when they only have textual inputs. That’s why so much work goes into training machines so they can make these types of intelligent decisions on the semantic level.
Ignoring this layered knowledge — i.e. which dishes are a similar cuisine and/or which dishes are made with similar ingredients — would lead to an unsatisfying user experience. Humans naturally create layers of associations between entities which leads to nuanced understanding of a concept. These relationships are simply too complex for a typical relational database.
So in order to make this real-world context machine-understandable, Uber built a food-based knowledge graph that showed the connections between restaurants, cuisines, and menu items. When users search for a particular type of food, Uber Eats can expand the search to include loosely related terms. What does this mean? Happier (and less hungry) users, and a more successful product.
Sure signs you need a knowledge graph
In addition to capturing real-world context, a knowledge graph effectively connects disparate data, creating a flexible data layer that adapts to changing requirements and changing realities. Just like you learn a new fact that adds context across many topics, a knowledge graph learns quickly too.
Whether re-platforming existing products or bringing a new offering to market, it can be hard to know what technology to choose. Today, global organizations are realizing that knowledge graphs play a critical role in solving a wide range of enterprise data management challenges by:
- Combining internal and external data: Instead of mastering external data to match internal definitions, the knowledge graph’s flexible data model easily connects related terms while allowing data owners to maintain control of source data.
- Supporting complex logical conditions and recommendations: When building a product with many conditional rules, a knowledge graph allows users to easily encode and centrally store all business logic, enabling powerful recommendation engines like those of Amazon and Pinterest.
- Addressing the need for fast release cycles: An inflexible infrastructure can hamper the delivery of feature requests or frequent content updates. By easily accepting new data, companies can quickly adapt to changing user and market expectations. A knowledge graph offers the flexibility to constantly update the knowledge base without requiring rework.
- Sharing an understanding of data across all devices: For IoT applications, a knowledge graph provides edge control and collaboration between proprietary systems and third-party systems/devices. The flexible data model offers quick connectivity of IoT devices for faster commissioning and changes.
- Connecting many formats of data: A knowledge graph can unify SQL, NoSQL, and unstructured data so that organizations can truly capture real-world context from all relevant sources. Knowledge graphs extract entities from unstructured data like full-text sources and connect those entities and their relationships into the knowledge graph.
The reality of digital transformation is that the majority of most “data-driven” efforts are doomed to fail. Data-driven efforts are doomed primarily because machines are not humans! Human decision-making is based on contextual intelligence, and in order to successfully automate, machines need to know what we know. Enterprise Knowledge Graphs (EKGs) have arisen to address this data management failure, bringing business meaning to machines.
Leveraging the inherent modern approach of EKGs, organizations can not only deliver new products that leverage their proprietary data, they can connect their internal data silos in a meaningful new way. Discover hidden facts and relationships through inferences that would otherwise be unable to catch on a large scale. Identify the nuanced meaning that different business units may have for the same entity. Create a data foundation, resilient to change, that keeps pace with continued shifts in the market. With business meaning captured in a machine-understandable format, companies can adapt for whatever comes next.
To learn more, visit stardog.com.
Kendall Clark is Founder & CEO of Stardog.
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. Content produced by our editorial team is never influenced by advertisers or sponsors in any way. For more information, contact [email protected]