Okay, You Got a Knowledge Graph Built with Semantic Technology… And Now What? |

In this article, we argue that a knowledge graph built with semantic technology improves the way enterprises operate in an interconnected world. With several examples, you will see how knowledge management can be made smarter using the potential of semantic technology to fuse data, analyze relationships, detect patterns and infer new facts from enriched datasets.

Frequently dubbed too difficult to build or too complicated to understand, semantic technology is increasingly gaining popularity. Again. In case you missed the party, travel back to 2015 or to 2006 or maybe to the moderately noisy one in 1999.

Whether you refer to the use of semantic technology as Linked Data technology or smart data management technology, these concepts boil down to connectivity. Connectivity in the sense of connecting data from different sources and assigning these data additional machine-readable meaning. Such an approach, no matter what name we use for it, is all about improving the way enterprises operate in an interconnected world. Examples of such continuous improvement are technological giants like Google and Amazon who use semantic technology principles to build better data architectures for better user experiences.

In these efforts to improve the way we connect with content, another term has found its way to co-occur (to use the industry-specific lingo) with semantic technology and it is the term knowledge graph. Sometimes related, and sometimes not, to the implementation of semantic technology, it is attracting everybody’s attention to the efficiency and the potential it brings to working with a lot of information.

Knowledge graphs are now also a buzzword among companies who are looking to integrate data from multiple sources and break the silos their legacy systems have left them with. As we read in Gartner’s report, knowledge graphs are “ ideally suited to storing data extracted from the analysis of unstructured sources “.

Not surprisingly, everyone seems to have or want to have a knowledge graph.

Why a knowledge graph?

Becoming more and more popular a term to denote an assemblage of technologies that help you find, manage and work with information, the knowledge graph built with semantic technology is attracting those who are interested in doing data right in the long-term.

Such knowledge graphs allow three extremely important knowledge management features, namely:

The above three power immersive interfaces, clear front-ends and, most importantly, enhanced content use.

The knowledge graph, glued by semantic technology, has the potential to merge all kinds of data, analyze relationships, detect patterns and infer new facts from the enriched datasets. For example, within a knowledge graph built with Linked Data, one can connect geographic, government, life sciences, commerce and many other types of data and easily explore, search, visualize and navigate the information they carry.

By ingesting and fusing data, the knowledge graph enhances the work of systems, apps, bots and search engines alike, giving this “machinery” (and us) the needed structured data to spin automation and increase the scale of the processes.

Take, for instance, the domain of business intelligence and the problem of discoverability. In this context, a big knowledge graph can help connect data and provide a context for integrating and analyzing company data aiming to facilitate structured information and, from there, enhance the discoverability of facts about business entities. (You can read more at: https://www.ontotext.com/knowledgehub/current/eubusinessgraph/)

In the healthcare industry, data integration is of paramount importance. There, internal company systems need to be harmonized and connected to external ones such as clinical decision support systems or automatic clinical trial matching. There is also a need for applying common clinical terminologies and standards. This is where a well-built knowledge graph can help break the productivity barriers and transform scattered raw data into structured knowledge. (You can read more at: https://www.ontotext.com/business-cases/medical-coding/)

Or, take the Pharma domain and the problem of mashing up information. A common issue there is how to get all, say, anti-neoplastic drug products for which patent protection expires in 2020, or how to find their known drug interactions within a specific drug category. Using a knowledge graph enables you to do that and you can arrive at deep insights by tracing such relationships. (You can read more at https://www.ontotext.com/drugbank-rdf/)

Okay, you got a graph, now what?

Weaving all kinds of data into one integrated whole gives you the tooling for different processes that aid knowledge work.

If you are an artist, like the author of the Poem Py, you could probably use the potential of a knowledge graph to level up the processes in your work. In this case, the artist scraped the text of an online poem and fed the words of the poem through the Google AdWords keyword planner. Then, he further returned the poem to a narrative order and format, added a checksum as ‘authorisation code’ and printed the monetized poem.

On an enterprise level, creativity and interconnecting worlds deliver targeted business value. One such example is when, in 2017 NuMedii commissioned Ontotext to build an expert knowledge graph with concepts from genomics, proteomics, metabolomics, disease conditions, drug products, scientific literature and various biomedical ontologies, integrated information from more than 20 open data sets. This massive integration helped the enterprise access highly normalized and semantically interlinked data, discover knowledge locked in documents and identify patterns and correlations between biomedical concepts.

Epilogue: Food For Thought

Any talk about graphs owes a mention of a blog piece by Tim Berners Lee’s blog called Giant Global Graph (archived). There, we can read about a Giant Global Graph and the many different people adding to the World’s knowledge. With more and more industries using semantic technology to build graphs of knowledge, it is not far-fetched to suppose that by building data spaces with semantic technology, the many different players using these for different reasons, will add to this Giant Global Graph. One knowledge graph at a time.

Teodora Petkova

Originally published at https://www.ontotext.com on July 26, 2019.

Providing a complete semantic platform for identifying meaning across unstructured data; Developer of GraphDB™, the industry leading RDF triplestore.