Assuming data are being analyzed from across the data landscape, commonly used analytical methods are not adequate to truly uncover important patterns and trends.
Data Fusion: Uncovering Data Patterns and Leveraging Multiple Data Sources
Data access and analysis is becoming increasingly important for both government agencies and commercial business alike. Most large distributed companies and government agencies store content and data in a variety of repositories distributed across multiple environments, systems and geographical locations.
Because it is generally difficult to conduct analysis across disparate data repositories, most firms and governments do not conduct comprehensive analysis across their data and information landscape.
It is easy to understand why this is true. In these types of environments, different data repositories almost always mean different types of storage methods, different naming conventions, and even different computing platforms. Together, this makes conducting analysis across these repositories difficult at best and impossible at worse.
Assuming data are being analyzed from across the data landscape, commonly used analytical methods are not adequate to truly uncover important patterns and trends. For example, it is very common that tabular reporting engines are run against data from multiple sources. Through these “traditional” reporting systems, a user can run queries against various data repositories and return results through a standard tabular output in the form of a table or perhaps even a pivot table. However, users are not given the power to investigate beyond a pre-defined number of relationships. In general terms, it is impossible to explore different levels of data, uncovering relationships and discovering unforeseen patterns in the data not easily detected by traditional means. In short, most businesses possess a wealth of data stored across multiple repositories but are not able to extract all the available underlying knowledge contained within those information resources.top