Obstacles in Information Integration

The basic obstacles in information integration across an enterprise are storage formats themselves, and the diversity of programmes that are used to process information. Typically, a representative talking to a customer has to extract relevant information from a variety of sources and formats, delaying response, and putting customer satisfaction at risk.

With strong information integration, these obstacles would be quickly overcome.
Naturally, there is a very strong movement towards such integration, and several companies are offering such integration.

What is on Offer

Ipedo, for example, has issued what they call an 'EII Challenge' where they get two days to integrate information given to them in a variety of formats: RDBMS, mail, web applications, XML, etc. Ipedo uses its XML Intelligence Platform for integrating.

This is an interesting offer, since traditionally there has been a segregation of structured data (mostly numerical), and unstructured data (mostly textual). When we talk about integration of unstructured and semi-structured information management into more structured information and numerical data we are talking of something like a complete Enterprise Information Integration or EII (a term supposedly coined by Aberdeen Group). Adopting this approach companies will be able to gain access to a 360-degree-view of complete organizational information - whether that information lies in databases in numerical format or lies in textual format somewhere hidden in a text field inside an RFP. EII - or something similar to EII - will be able to offer an overarching information-umbrella to integrate: Corporate Information Portals and Enterprise Content Management Systems; Enterprise Applications: CRM, ERP and SCM; Business Intelligence and Decision Support Applications; Web services; and, Unstructured and Semi-structured Information Management Platforms.

We should also remember that structured and unstructured data also meant two different domains. Several companies that processed unstructured data were fond of pointing out that almost 80 percent of data in an enterprise is actually unstructured data.

IBM, however, has approached the problem from what might perhaps be seen as an unexpected angle: it is offering to integrate unstructured data into its database programmes. Code named 'Masala' a beta version; this product will integrate and process a broad range of content types: email, spreadsheets, XML, text files, voicemail etc. It is supposed to have a hundred new features. This approach too is an interesting advance, though at the beta stage.

Convergence?

Perhaps the current buzzword, convergence, is operative here too. A true convergence of way of treating structured and unstructured data will in the end only be beneficial to all. Instead of segmenting various processes, and dividing the market into zealously protected turf-ending in turf-wars-a convergence of technologies will enable various companies to co-ordinate overall development, collaborate, and hand down more benefits to end-users.

Convergence will also further, in a fundamental way, the attempts to reach that elusive goal of standardization-but from a different angle altogether-through flexile and versatile software, rather than reluctant compliance.