Searching for the integration solution

We are now well into the season of surveys, but here is one that begs an interesting question – is application integration as important as it is usually assumed to be?

It is widely accepted that the way to build collaborative applications and services is to use integration tools to bind them together, or at least let them exchange relevant data in a useful manner. Yet the results of this, according to research conducted on behalf of semantic search company, Ontology Systems, can be less productive than many businesses might have hoped.

According to the research findings a whopping 83percent of IT staff do not expect to see any return on their investment from current integration projects over the year. Set against this, however, it also shows that 84 percent believe that better use of data, usually the prime reason for integration projects in the first place, could lead to improvement in customer service.

The research, the company suggests, indicates that businesses see this as a treadmill, that is also their only option when trying to innovate new ways of exploiting business data and gaining a competitive advantage. As a result IT managers are allocating as much as 25 percent of IT budgets into integration projects, many of which can be both expensive and time-consuming. Yet, the research shows, some 65 percent of respondents don’t ever expect to see a return on the effort and investment.

Needless to say, Ontology also suggests that there is now an alternative to application integration. What is more, it is one that can be made available as a SaaS, run as a hosted cloud service, or installed on-premise.

 “Data should be being used to gain a competitive advantage, not a hindrance, “says Benedict Enweani, CEO, Ontology Systems. “In theory, integrating systems would be the most effective way of getting a unified view of data held across a business. However, most real world systems were never intended for this approach, which means that critical business assets can’t easily be found when their associated data spans more than one database or system, as stated by 56 percent of respondents. This results in not only a costly, but also a very time intensive procedure, as systems take a long time to merge and integrate.”

Ontology’s solution is semantic searching, a much more complex process than a typical, index-oriented searching process. This can search through multiple datasets using a variety of complex concepts, such as the context of search, the location of the data, inference of the searcher’s intent, variation of words, synonyms, generalised and specialised queries, concept matching and natural language queries. Combinations of any and all of these can be brought together to identify relevant search results that can then be exploited.

It has the advantage of being faster to implement than integration processes, and far more flexible, in that it can search all the available data in the enterprise, or any of it that is specified for the search.

It is possible, in fact, to see semantic search as a complement to integration as well as an alternative. It could readily be used to identify those applications which could provide real business advantage by being integrated, as well as a source of integrated data in its own right.

It can also be used for searches with far wider implications. For example, it can also used to identify security or operational issues which might be harmful to the business. Such issues could range from identifying poor management of business processes which could cause compliance and governance issues, through to detecting attempts at fraud.

That flexibility of approach, coupled to the fact that it can be made available ad both SaaS and as a hosted service, make it particularly useful for operations with a large cloud- or web-based component.

As Enweani concludes; “You wouldn’t dream of integrating the internet. Google revolutionised access to online data when they enhanced Internet search. To gain true business advantage organisations need to take the same approach. When it comes to enterprise data, we say ‘Search don’t Integrate’.

“The majority of today’s companies are working towards business goals that depend on the efficient use of their data estates. By using semantic search, businesses can completely side step the traditional data integration approaches that are failing to help them extract maximum value from their data and allow them to find, examine and link data held in sources right across the business and get the most from their data.”