Considering that ontology’s primary purpose is to represent knowledge within a domain, it could offer an alternative to pre-packaged solutions to enterprise data issues. Primarily, an ontology, which is a representative semantic model, is intended to convey shared understanding within a domain, often via the implementation of a knowledge graph, such as a triple store. Moreover, it emphasizes the relationships and meaning of those relationships between classes within a business.
Traditional data modeling is largely confined to the capture and retrieval of data. In contrast, an ontology concerns itself with a shared understanding of what that data means. Hence, committing to enterprise ontology forces organizations to understand and document the domain for which they intend applying natural language processing and machine learning. Moreover, a semantic data model integrates conceptual, logical and physical models along with a deep data dictionary and glossary into the ontology to facilitate business understanding. Furthermore, the flexibility and power offered by linked data, using open standards and a handful of mostly open-source tools, is immense.
The advantages of semantic data models and the use of triple-stores or other types of graphs) are well documented. The two most obvious reasons are, however, linked data and using inference. Reasoning is difficult to perform in relational databases and can be computationally expensive, whereas ontologies excel at reasoning. Also, it is relatively easy to develop algorithms that explore graphs for new relationships that had not been defined in the model.
Semantic modeling is concerned with meaning, and hence new relationships can provide insights within the context of a domain. This logical reasoning is in contrast to the statistical approach often taken with data-lakes or traditional natural language processing, which might be good at finding similar concepts, but miss different concepts that are relevant or actually share meaning.
Committing to an enterprise ontology, linked data, and open standards allow organizations to control where and how to leverage people and capital. In the longer term, an enterprise ontology gives organizations the freedom to implement own ideas, rather than waiting for them to become commoditized and packaged by big-box vendors.
To summarize, an ontology might be the answer to building an enterprise knowledge base, supported by a robust representational data model the will provide the building blocks to be self-determinate, have an insightful view into data and not be beholden to someone else’s ability to innovate.
Click here to read the original article published in LinkedIn.
Please give your feedback on this article or share a similar story for publishing by clicking here.
