Business Intelligence solutions can inherently be Data Integration Solutions. To begin with, there must be access to a wide array of relevant data. There must be an effective process for integrating data across different sources; if data is missing from the analysis the results derived will either be incomplete or inaccurate. Data integration solutions require the marrying of the right architecture with the right processes based on the specific requirements of the business. But within the BI community, there is no real consensus on what is the most effective means of data integration. What are the best data integration solutions?
For a long time data warehousing was the primary architecture on which BI solutions were built. These solutions were geared mainly towards big businesses or enterprises where there is a large volume of both current and historical data (not to mention a large budget!). Data is stored & structured within the data warehouse or between data marts to facilitate efficient data retrieval. However, while solving problems relating to the dispersion of data across different locations and problems with data quality, data warehouses introduced a new set of challenges. Data warehouse models proved to be very costly and resource dependent often backed by performance and data consistency challenges. In fact, whereas data warehouses aim to integrate data into a centralised source, it oftentimes becomes a problem in itself as data is dispersed across data warehouses and data marts. Using data warehouses and associated ETL approaches as data integration solutions is often the most costly approach. It’s often not the best data integration solution. Certainly, other data integration solutions should be considered.
New approaches to data integration are constantly evolving and data mashup now promises a unique advantage. BI solutions modeled on this approach even lend themselves to SMBs / SMEs. The data mashup approach to data integration does not set out to centralise data sources, but rather once the source of the data is identified allows the data to be integrated on the fly. There are many advantages to integrating data using a mash up approach as performance, data integrity and consistency issues can often be averted. Furthermore, data integration solutions using a mash up approach proves much more cost effective and demands far less resources. While BI solutions modeled on a mash up approach to data integration can operate in the smaller business sectors, these solutions often have the capacity to sit horizontally on existing applications, including the existing data warehouses that are often in place at large enterprises. So it’s very scalable. Add to this the ability of data mashups to integrate both internal and external data … and that’s why we think the data mashup approach is one of the best approaches to data integration solutions.
Recent Comments