A Hierarchy of Metadata Elements for Business Intelligence Information Resource Retrieval

Introduction And Motivation

Information is a strategic asset for today’s organizations. The ability to effectively capture, organize and exploit information resources can provide a competitive advantage, and is critical to today’s information and technology rich organizations (Zheng, Fader & Padmanabhan, 2012). Organizations build business intelligence (BI) applications to meet the information needs of knowledge workers (Ramakrishnan, Jones, & Sidorova, 2012; Kiron, Shockley, Kruschwitz, Finch & Haydock, 2012). Business intelligence applications provide users with direct ‘ad hoc’ access to structured data, as well as to prebuilt reports, scorecards and dashboards based on this data (Chaudhuri, Dayal & Narasayya, 2011). In addition, it is common in organizations for individuals to export data generated from business intelligence applications (and other data sources) into structured documents (such as spreadsheets) for further analysis and manipulation and to create unstructured documents (such as presentations and documents in tools such as Microsoft PowerPoint and Word) that analyze and interpret this data (Roberts, 2010).

As a result, there is currently an ever-expanding pool of heterogeneous information resources that need to be effectively managed in order to be reusable and by extension, valuable to the organization. The need for large volumes of information has resulted in challenges for many organizations as an overwhelming volume of data, reports and documents are created on an ongoing basis (Davenport & Snabe, 2011). In a seeming paradox, the more information that gets created in an organization, the less it is used, as users and potential users become overwhelmed by the sheer volume (Bruce, 2005). Addressing this issue requires an effective information retrieval mechanism: a powerful and easy to use method that allows individuals in an organization to search for and retrieve the information resources they need. However, modern information technology (IT) departments are struggling with how best to identify information resources and to organize and catalog them in a way that they can easily be retrieved. An effective search mechanism requires the right metadata (Efron, 2011). Metadata can be thought of as data ‘tags’ associated with an information resource that describe the characteristics of the resource (Hurst, 2010).

This study seeks to determine the feasibility of employing a standard set of metadata elements to aid in the retrieval of three different BI-oriented information resources, specifically: (1) structured BI generated reports; (2) structured reports implemented in tools such as spreadsheets; and (3) unstructured reports from Word, PowerPoint, and similar applications. The objectives of this study are to:

• 1.

  Develop a hierarchy of metadata elements used by knowledge workers when searching for structured-BI, structured non-BI, and unstructured information resources;

• 2.

  Identify the most (and least) useful metadata elements for information retrieval in a Business intelligence context by knowledge workers; and

• 3.

  Explore how the relative usefulness of metadata elements to knowledge workers vary on the type of information resource (i.e., structured-BI, structured non-BI, unstructured) to be retrieved.