Read the first post in our new series of blog posts focusing on major Ontotext’s webinars and how they fit in the bigger picture of…
In the last few posts dedicated to Ontotext webinars we have introduced some of the partners comprising our rich technology ecosystem. This partner ecosystem powers the next generation of Content and Data Management applications for many of the world’s most knowledge-intensive enterprises from Life Sciences and Financial Services to Publishing, Government and Industry. The result: the ability to effectively meet the requirements and provide the full capabilities of end-to-end enterprise solutions – something no single vendor can do on their own.
Today we want to introduce another partner – Enterprise Knowledge (EK).
EK is the largest consulting company in the world dedicated to knowledge management and, as of 2021, they have been recognized as one of the KMWorld 100 Companies That Matter in Knowledge Management for 6 years in a row.
Ontotext is happy to partner with EK because of their mature and competent approach to understanding customer’s needs and smoothly driving the process of defining and delivering a sustainable solution. At the same time, EK has a long bench of competent architects and engineers with experience in semantic technologies, content and data management.
Historically, knowledge management has been mostly about controlled vocabularies and taxonomies, which had the sole purpose of helping manual document indexing. This is no longer the case. More and more, taxonomies play an important part in various content management and data management systems.
In today’s competitive environment, enterprises have to make decisions based on knowledge that resides in both structured and unstructured information coming from disparate internal and external sources. Not surprisingly, leading organizations increasingly rely on big teams of data scientists. But to get a real advantage, these teams need to work with rich and high-quality data. They need to get the right features and signals on time, without massive efforts for ad hoc data cleaning, transformation and normalization.
It’s true that today’s data warehouses and enterprise search systems manage well when it comes to structural transformations and indexing. But they struggle with aligning the meaning of the information on all the necessary levels: terminology, data schema and individual entities.
Read on and watch the webinars to get to the details!
Our first webinar Knowledge Management Becomes Business Critical as Knowledge Graphs Improve Decision Making and Efficiency is presented by Atanas Kiryakov, CEO of Ontotext, and Joseph Hilger, COO of Enterprise Knowledge, and focuses on how knowledge graphs not only align with knowledge management but also make it critical for an organization.
You learn how knowledge graphs address knowledge management challenges by making concepts and entities first-class citizens. Interlinking a critical mass of domain knowledge, which is richly described with unambiguous formal semantics, brings context. This brings a couple of very important benefits:
In this webinar, Atanas introduces a mindmap of knowledge graph applications and demonstrates how Ontotext Platform supports the different stakeholders. Then, Joe talks about EK’s basic methodology for knowledge graph centric solutions. He explains how their hybrid analysis approach, which blends user- and technology-driven research, enables them to iteratively capture relationships between the application, business, information and technology layers. He also presents several use cases about Enterprise Architectures, Natural Language Search, Recommendation systems, Data Portals and Catalogues.
Interesting attendee question: How may I present the knowledge graph approach to the C-level management of my organization?
Ontotext and EK answer: The key is to find a specific solution that would make a difference. Sometimes the issue is that people don’t know what they need to know in a timely fashion. Or they need to democratize access to data so they don’t have to go to IT to solve a problem. In any case, if you go to your executives with the mindmap we just presented, they would be scared.
The best way to pitch the knowledge graph approach to your executives is to figure out what are the specific things that can be used to solve a specific problem your organization faces. Then you can go and say: “Here is a problem we all know we have and I found a way to solve it. And, oh, by the way, it can also solve some other problems we have.” We’ve had a lot of success starting with one project and then doing three, four, five more after that because people just saw the value.
The first webinar focused on the role of knowledge graphs in making knowledge management an essential business function. Our second webinar Turning a Taxonomy into a Recommendation Engine delves into the development of content recommenders using taxonomies and GraphDB as a foundation for building enterprise knowledge graphs. It is presented by Connor Vilenio, Senior Data Management Consultant at EK.
Taxonomies are hierarchical metadata structures commonly used in knowledge management to capture the context, meaning and business value. As huge amounts of the information created by enterprises is unstructured (free-text documents, images, videos, etc.), which makes it difficult for machines to process, taxonomies help organize this unstructured information and better describe content.
In this webinar, Connor explains how taxonomies present a great starting point for building knowledge graphs that support personalization services. He shares his experience with improving the underlying graph model of a taxonomy to support a content recommendation system. This helps create better user experience by offering relevant content from a semantic model, which reflects the enterprise’s understanding of semantic relevancy.
You learn about some common use cases that benefit from a recommender system and how taxonomies can help calculate relevancy. You also find out how an enterprise can get meaningful features from the structure of their graph, which helps model and then calculate semantic relevance. Finally, you learn how, by using GraphDB as a foundation for driving content recommendations, you can enrich taxonomies with custom relationships and polyhierarchy, and turn them into knowledge graphs.
Ontotext and EK answer: It all goes back to how the project was designed in the beginning. In other words, what’s the alignment between the taxonomies that the enterprise already has and any potential ontology projects that are going on. The reason we find taxonomies useful is that – because, typically, taxonomies already have use cases around organizing and adding meaning to content – when you want to make, for example, content recommenders, this is a great place to start.
That being said, it’s not just your taxonomy that goes into the model, but it’s usually expanded with other ontology properties and structures. So, you can either do it from scratch and develop a new ontology or maybe there is an existing ontology that can be useful – either open source and publicly available or something that is already being developed in the enterprise. Again, the main thing is what the alignment of your use case is and then you use your project design as a North Star to decide what is the best starting point.
That covers the two webinars that we wanted to present to you today.
To sum it up, our extensive partner technology ecosystem effectively covers the capabilities and requirements of end-to-end enterprise solutions. The blend of our technologies, expertise and consulting capacity allows us to deliver knowledge graph centric systems which:
We hope you’ve heard enough to want to dive straight into one of them or why not both?
Watch out for our next post in this series At Center Stage or visit our Webinars directly to check out what is upcoming!