In addition, the support of multiple taxonomies is also critical for a data warehouse, and to the extent the architects have created a database architecture that will provide for metadata definition and re-defining of taxonomies is the extent to which the data warehouse will have greater use in the organization. Without a strong focus on these aspects of data agility, a data warehouse can quickly become outmoded and only marginally successful. Assume that you are the data quality expert on the data warehouse project team for a large financial institution with many legacy systems dating back to the 1970s. Review the types of data quality problems you are likely to have and make suggestions on how to deal with those.

There are going to be a myriad of data quality problems inherent in managing the data quality inconsistencies with legacy systems dating back from the 1970s. Most significant and potentially problematic are going to be the byte-ordering inconsistencies from operating systems during that era which significantly influence how portable the data between systems can be. As a result, there is often are inconsistent data values to the byte order level that must be resolved through specialization translation applications. On conjunction with this shortcoming is the inconsistent and incorrect data formatting that is inherently included in the data. Second, there are often entities, interrelationships...

Third, there is the difficulty of how to take into account the significantly different timeframes of each dataset, as one may be based on entirely different set of assumptions than others. For example one database could be specifically based on a timeframe of six-month data while another could be based on over a daily recording of transactions. This would require significant redefinition of the data to make it useable. There are also the issues of which database encapsulation approaches were created for the original data and how that is reflected in the overall dataset as well. In addition to all these other aspects there are the issues of having applications that have fragmented data sources, inaccessible data due to a lack of semantic consistency across all systems and a lack of consistency of how the object model has been specifically designed for the data as well, on top of all these issues there is the most significant, and that is modifying how the organization will use the revised data set once put into place for re-integration to the applications. Change management at the application use level is one of the most daunting tasks as people often do not want to change how they do their daily jobs.