Management Support with Structured an Unstructured Data - An Integrated Business Intelligence Framework

Baars, Henning, and Hans-George Kemper. Management Support with Structured and Unstructured Data-An Integrated Business Intelligence Framework. Information Systems Management 25, no. 2 (March 2008): 132—148.

Introduction

Three different trends have been drivers for the increased importance of BI

1. more turbulent, global business environments
2. pressure to unveil valid risk and performance indicators to stakeholders, and
3. aggravated challenges of effectiveely managing more and more densely interwoven processes.

The most salient BI tools for addressing these challenges are still:

1. reporting
2. data mining and
3. OLAP

Business Intelligence Framework

BI frameworks provide a structure for BI infrastructure that enables holistic decision support. In the past there have been three classes of BI frameworks building either on

1. the concept of the data warehouse focusing on the technical aspects
2. a broad organizational and demand-driven view
3. partial structures for specific approaches.

Baars and Kemper introduce a Multi-Layer BI Framework that consists of the following layers:

1. data layer - stores structured and unstructured data which has usually been transformed using Extract Transform Load (ETL) components.
2. logical layer - responsible for data analysis and includes processes such as OLAP, data mining and the generation of business reports and ad hoc analysis.
3. access layer - provides easy access to these functionalities, e.g. by using some kind of portal software.

They also distinguish the following approaches for integrating structured and unstructured data

1. integrated presentation - the integration takes place in the access layer only
2. analysis and content collection - unstructured content is already handled at the data collection layer
3. distribution of analyses results and analysis templates - connects components on the logic layer (e.g. analysis systems and components for knowledge distribution such as a CMS); automated processes may extract relevant business data from the CMS and automatically publish documents summarizing the result of BI processes.

Applications

Customer Relationship Management (CRM)

CRM binds all customer oriented activities together and, therefore, integrates all marketing, sales, and service processes:

1. Operational CRM: supports day-to-day activities
2. Collaborative CRM: integrates the management of all customer touch points (mail, phone, ..)
3. Analytical CRM: performs data analysis

Competitive Intelligence (CI)

CI focuses on all processes for gathering and analyzing information about the competition and the general market environment (SCIP, 2005).

1. the most relevant CI information is from outside the company borders (financial reports of competitors, government publications, patent databases, Web, .... (compare Lux and Peske, 2002)
2. Examples:
• analysis of press reports reveals weak signals regarding shifts in public opinion
• product reviews indicate customer satisfaction
• patent analysis using clustering yields research trends and the strategic orientation of the competition; estimate how this might affect the future sales of certain product groups
• market data might reveal the impact of new substitute produces on sales