As part of a long-term study for the European Commission, Ovum has conducted in-depth interviews with IT managers to determine the drivers for and barriers to adopting semantic software within the enterprise. While there is some resistance to the term semantic, a more substantial barrier brings us right back to basics: finding employees who know the business. Vendors and implementers of semantic software should take note.
Resistance to semantics and new technologies
The Value-IT project, part of the ECs Seventh Framework research programme, is assessing the potential demand for semantic technologies in the enterprise (STEs). Ovums analysis experience is being used to identify key issues throughout the two-year study period.
Interviews to date clearly indicate that semantic software is growing in popularity. This label covers a range of applications which understand meaning and context of information within a business, and can therefore be more effective than coding in joining data, systems and processes together to serve both employees and the business as a whole.
However, various preconceptions about technology often spring from the technology being launched prematurely or bandied too early, slowing uptake when it finally matures. Ovum has taken a close look at the use of the word semantic, asking IT managers whether they believe it carries negative baggage.
Overall, the word is not a severe hindrance to the development of STEs, but does alienate some people. As one interviewee explained [the term] semantic web is a barrier. Its associated with a vision of the web for ten years to come. I make a point of not using [the term] semantic web.
However, some interviewees felt that, while the term semantic may have been overused a few years ago, technology has now progressed to the point of legitimacy.
The real barrier: finding people that know the business
A more critical barrier to the adoption of semantic technologies is an old-fashioned one: finding people that know the business. Generally, in order to implement semantic technologies, organisations, processes and associated data/terminology must be mapped out into a set of binding concepts, often embodied in an ontology.
However, it is hard to find employees who possess both a breadth of oversight across various departments within a company and in-depth knowledge of the processes and terminology within those departments. Therefore, a significant challenge lies simply in finding a person expert enough to populate an overarching scheme or ontology which would map out individual processes and terms from different departments.
One interviewee specifically underlined this issue. My biggest concern is that we wouldn't have the resources and find the right skills to deal with such kind of information […] because you have to define and describe something, and how things should work together, and how they should be combined. To do this is difficult and the risk is we wouldn't have the right people.
Sign up for Computerworld eNewsletters.