What makes a business ripe to adopt semantic web technologies? Those engaged in cross-enterprise business processes, in particular where models based on web technologies drive greater collaboration and increased dynamism, are on the list, says Professor Adrian Paschke, Corporate Semantic Web chair at the institute of computer science at the Freie Universität Berlin and head of the InnoProfile project Corporate Semantic Web.
“That is motivation to apply semantic web technologies because you no longer are working in closed walls where you build your own schema and database model, but you need a flexible semantic model that easily integrates with others,” says Paschke.
You need, that is, Linked Data to open up and integrate closed repositories and a semantic layer to interlink ontologies.
Paschke will be speaking about bringing semantic web technologies to the enterprise at the upcoming Semantic Tech & Business Conference in Berlin. Those companies that would do well to invest in semantics also include any whose business models revolve around information and knowledge – media, which is coming more to that understanding, for example, or life sciences, which was early to grasp it.
Such businesses can benefit from lessons learned by those who went before them. One point, for instance, is to skip the top-down approach that many initially took, with its focus on designing ontologies and painstaking modeling accuracy, which takes a lot of time and whose outcomes may be unknown.
Rather, Paschke says, “particularly in the corporate context, start with existing data and existing schema and try to translate that to the semantic web. …It’s easier to start bottom-up with data and transforming it with automated tools, and [move forward once] there are more semantics in the data.”
That seems in line with enterprises becoming more practical about applying semantic web technologies now that they have a better understanding of its contribution to transforming information that has been translated from data into actual knowledge. A year or two ago, he says he heard more questions from people in business about how semantics applied in a corporate context, whereas now it’s more about issues such as where the costs are, the ease of use of tools, and how much training and experience employees need to create ontologies.
Speaking of economic considerations, those extend to semantic web champions making a solid case to decision-makers around cost models. “Often there is an existing CMS or document management system, and when there is a decision about replacing that with a semantically-enriched system, you’ve got to be able to demonstrate why,” he says.
Demonstrations that can show how having such a semantic system in place leads the way to faster decision-making or higher output for a department, and the economic impact of that, are going to become important. Cost-model and –impact examples for things like ontology engineering aren’t exactly widespread yet, though.
And champions need to create incentives for the business users to actually use the semantic web systems to be developed, since at least initially those people will need to make some extra effort to put additional metadata into systems. They’re going to need strong reasons about how benefits will accrue to them from doing so, or they likely won’t do it, he says.
Another piece of advice he has for the enterprise with semantic aspirations is to smoothly integrate semantic web technologies into existing systems, so that users aren’t being asked to undertake a learning curve both of using semantic tools and completely new interfaces, too.
To register for SemTech Berlin, visit this page.