Ontology's Five Principles

Ontology delivers solutions quickly because its semantic search and modelling capabilities are built around proven graph-based technology featuring five differentiating principles.

Ontology was built to address the problems that cause delay and risk in traditional data integration projects. Using the power of semantic search and built upon proven, graph-based technology, Ontology was designed around five differentiating principles: No Schema, No Integration, No Big Bang, No Search Restrictions, No Upfront Risk.

No Schema

Ontology’s flexible, model-based approach and use of graph technology makes it possible to combine business entities from multiple, disparate data sources into a semantic model without rigid schemas. Ontology can incrementally model relationships and parameters of business entities and, unlike traditional integration, easily accommodates changes in requirements. The resultant semantic model understands the business entities, entity relationships and the dependencies across different entities in the data. 

No Integration

Ontology doesn’t focus on integrating the schemas, meta-data or the structure of data sources. These artificial, technical constructs only serve to hide business entities within a rigid, yet fragile structure. Instead, Ontology works from the data, finding and combining business entities fragmented across different enterprise application data sources. This speeds up the process of gaining access to meaningful information within data from multiple sources.

No Big Bang

Ontology’s “just enough modelling” approach means that only the business entities that are required for a project are modelled. If more entities are needed, they can be modelled from existing data sources. If additional data sources are added, they are simply accommodated by finding and combining the relevant embedded business entities, thus building more intelligence into the semantic model over time. The result is that as project requirements change, the Ontology model of the data simply evolves to accept the changes without the need to manage the process of unifying and federating rigid data schemas. Ontology delivers value early and iteratively over the life of a project.

Ontology provides a choice to organisations about how much they want to invest to create the requisite intelligence in their data. With little or no modelling of the combined enterprise application data, low cost search results can be presented that rely on the user’s domain knowledge to search and surf to find the information they need. Over time, the intelligence in the model can be incrementally increased by modelling more of the data, creating more data transformation rules and designing more presentation views. Each such increment increases the intelligence in the model and produces increasingly more accurate, more automated and more valuable search results. By providing the choice to invest incrementally in the intelligence in the semantic model, Ontology allows organisations to lower the cost per datum of turning data into meaningful, accessible, exploitable knowledge.

No Search Restrictions

In a schema-based world, if the data isn’t in the schema, it has no search value. Ontology loads or retains access to all data sources whether it is later modelled or not. This means that all data from the combined sources can be searched. If new data is added, this also becomes available to be searched. If user requirements change over time, and access to other business entities is required, these can simply be modelled incrementally, adding more and more search value to the model.

For organisations that need to understand everything they ‘know’ about a business entity, for compliance or other legislative reasons, it is not enough to be able to just look across the application data for this information. Ontology allows business entity information found within enterprise application data to be linked to every mention of that entity within the organisation’s unstructured data – its documents, spreadsheets, PDFs and emails.

No Upfront Risk

We looked at the areas of technical and commercial risk within data access projects, and created a combination of product features and commercial policies that, as far as is possible, mitigate this risk. 

Project feasibility. Understanding whether a project is technically feasible can be understood within 2 weeks with Ontology. This is because, from day one, Ontology works with the data and not from an arbitrary schema to build relationships and linkages between data sources. This means that if it is not possible to infer a linkage between the same entity in two data sources, this is rapidly clear – because the data is just not there. With traditional data integration solutions that must first model the schemas to support the integrated data, the fact that data sources cannot be linked may not become apparent until data is loaded into the schemas. 

No capital expenditure and no long-term commitment. An Ontology solution can be implemented using operational expenditure and requires no lengthy, contractual commitment.

No hardware required. An Ontology project can run at your own premises, or if preferred it can be hosted by Ontology. This can remove the capital expenditure and IT provisioning barriers that can otherwise introduce unforeseen project delays.

No unnecessary tying up of your team’s resources. As team resource is precious, our experts are versed at building business semantic models from disparate data sources with very low involvement from your team. 

Data availability. Working with IT to build links to access data in live systems can be a technical, as well as a political challenge that can delay projects. Ontology needs nothing more than a recent extract of the data. 

Changes in requirements. Ontology’s inherent support for incremental project methodologies like Agile, coupled with a semantic model that evolves to accommodate changes in requirements, allows organisations to embrace project changes. Data sources, business entities and search needs can be changed at any time in the development of the solution, and the model simply evolves to accommodate the changes. This reduces the chances of project overruns, or of delivering projects that don’t fully meet requirements.

The Result

Ontology has adopted these these five differentiating principles into its technical design, project approach and commercial policies. The result is a completely new approach to solving data access challenges that significantly reduces the timescales, cost and risk associated with traditional data integration projects.

Search, don’t integrate.