Authors:
Emiel Caron
1
and
Hennie Daniels
2
Affiliations:
1
Erasmus University Rotterdam, ERIM Institute of Advanced Management Studies, Netherlands
;
2
Erasmus University Rotterdam, ERIM Institute of Advanced Management Studies; Center for Economic Research, Tilburg University, Netherlands
Keyword(s):
Business Intelligence, Multi-dimensional databases, OLAP, Explanation, Sensitivity analysis.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Business Analytics
;
Communication and Software Technologies and Architectures
;
Computational Intelligence
;
Data Engineering
;
Data Warehouses and Data Mining
;
e-Business
;
Enterprise Information Systems
;
Enterprise Software Technologies
;
Evolutionary Computing
;
Information Systems Analysis and Specification
;
Intelligent Problem Solving
;
Knowledge Discovery and Information Retrieval
;
Knowledge Engineering and Ontology Development
;
Knowledge-Based Systems
;
Machine Learning
;
Management Information Systems
;
Soft Computing
;
Software Engineering
;
Symbolic Systems
Abstract:
In this paper, we describe extensions to the OnLine Analytical Processing (OLAP) framework for business analysis. This paper is part of our continued work on extending multi-dimensional databases with novel functionality for diagnostic support and sensitivity analysis. Diagnostic support offers the manager the possibility to automatically generate explanations for exceptional cell values in an OLAP database. This functionality can be built into conventional OLAP databases using a generic explanation formalism, which supports the work of managers in diagnostic processes. The objective is the identification of specific knowledge structures and reasoning methods required to construct computerized explanations from multi-dimensional data and business models. Moreover, we study the consistency and solvability of OLAP systems. These issues are important for sensitivity analysis in OLAP databases. Often the analyst wants to know how some aggregated variable in the cube would have been chang
ed if a certain underlying variable is increased ceteris paribus (c.p.) with one extra unit or one percent in the business model or dimension hierarchy. For such analysis it is important that the system of OLAP aggregations remains consistent after a change is induced in some variable. For instance, missing data, dependency relations, and the presence of non-linear relations in the business model can cause a system to become inconsistent.
(More)