Making Progress on Centralizing Data Analysis

In a column for this month's Waters magazine, an analysis of data management culture and methods touched on the idea of centralizing data tasks. I've been hearing a lot more since then about how data analysis is a function that can benefit from centralization and consolidation of data.

Generally, when computers are given multi-dimensional analytical inquiries, they use online analytical processing (OLAP) as a means to produce answers to those questions. With new and more complex middleware, visualization technology and lower latency messaging all in the mix, a new version of OLAP could be needed to include these technologies in processing and answering those inquiries.

There are some technologies that can extend data models rapidly and on an impromptu basis, particularly for the definition of a security. Firms are starting to use centralized data groups to validate models. Such centralization was once considered rigid and inflexible, but now with rapid application development (RAD) environments, even a centralized environment can be flexible or be altered to keep pace with changes in data due to business changes.

More lightweight applications, cloud computing and the use of big data can all contribute to centralization of data analysis. While cloud computing can provide value by making data sets available for different analyses in different applications, and also decrease the time required to do different analyses, connecting different clouds together, and connecting internal data sources with data hosted using cloud resources, can present a challenge.

Still, it does seem that centralizing data analysis is not the unwieldy proposition it once was, and firms are implementing it, although with some caveats. There is room for advances and improvements, and there will be more developments to come.