Towards Ad-Hoc and Collaborative Business Intelligence

Introduction

Current Business Intelligence (BI) environments suffer from several shortcomings regarding the following characteristics:

• • Focus on individual needs: Today, business users cannot fulfill their individual information needs but have to rely on standard reporting and predefined analytical content; they depend on either IT administration or enhanced technical skills to generate business reports.

• • Availability of business context information: Highly relevant business context information, such as definitions, business goals and strategies that help to understand the results of an analysis are missing in current BI systems as well.

• • Support for collaboration: Despite the fact that decisions making is a collaborative task, today’s BI solutions only provide no or rather fundamental collaboration capabilities.

• • Open Business Intelligence world: Today’s BI solutions have a strong focus on structured, internal information but lack of integrating external and/or unstructured information.

Failing to meet the BI challenges described above inevitably leads to poor usage of BI systems. According to the world’s largest BI study of the Business Application Research Center (BARC, 2009), in only 23% of all enterprises more than 20% of the employees use BI technologies; the overall mean ratio of employees using BI regularly is just 13.2%.

We envision a BI platform that allows business users to derive their own business information in an ad-hoc fashion, discuss the gained knowledge and shape their business strategies in a collaborative manner. This platform will reduce IT dependencies and put information acquisition directly into the business user's hands, be it managers or operative information workers, see Figure 1.

Figure 1.

Envisioned platform

We propose techniques to significantly enhance the individual steps along the BI process. These techniques ease the business configuration and improve the capabilities to find the relevant information and to analyze the data; we make collaboration an essential part of the decision making process and enable the tracking of taken decisions. Further, the platform may utilize a variety of different data sources, which clearly positively impacts the analysis step. To accomplish these goals we need:

• • A flexible data model to describe and adapt business relevant entities and their relationships within and across organizations. Such a data model allows to react on rapidly changing conditions by adding, removing or modifying new entities and relationships. Further, it allows to model various context information (i.e. information relevant for the decision making process that complements the data to be analyzed).

• • An efficient and scalable data store for the business data to realize fast response times even for large datasets. Particularly, this data store has to take the flexible data model into account.

• • A business configuration methodology to populate the data store in a pay-as-you-go fashion and to empower users to adapt the business data on their own.

• • An information self-service allowing users to easily retrieve the business information of interest. The information self-service should enable business users to satisfy their information needs in an intuitive, fast, and efficient manner and to define a personalized environment for retrieving information of central data sources, integrating external data sources, and analyzing both kinds of information in an integrated manner.

• • An integrated collaboration environment to discuss the results, derive conclusions and take decisions in a collaborative fashion. Important properties of this environment are synchrony, intuitiveness and traceability.