2001 — Seventh Americas Conference on Information Systems 329
DEVELOPING REQUIREMENTS FOR
DATA WAREHOUSE SYSTEMS WITH USE CASES
Robert M. Bruckner
Vienna University of
Technology
Beate List
Vienna University of
Technology
Josef Schiefer
IBM Watson Research Center
Abstract
Intelligent and comprehensive data warehouse systems are a powerful instrument for organizations to analyze
their business. The implementation of such decision systems for an enterprise-wide management and decision
support can be very different from traditional software implementations. Because data warehouse systems are
strongly data-driven, the development process is highly dependent on its underlying data, which is generally
stored in a data warehouse. Since data warehouse systems concern many organizational units, the collection
of unambiguous, complete, verifiable, consistent and usable requirements can be a very difficult task. Use cases
are considered as standard notation for object-oriented requirement modeling. In this paper we show how use
cases can enhance communication between stakeholders, domain experts, data warehouse designers and other
professionals with diverse backgrounds. We introduce and discuss three different abstraction levels (business,
user and system requirements) of data warehouse requirements and show how use cases can be drivers for the
requirements development.
Keywords: Data warehousing, requirements engineering, use case modeling
Introduction
Building a data warehouse is a very challenging task because it can often involve many organizational units of a company. A data
warehouse is a common queryable source of data for analysis purposes, which is primarily used as support for decision processes.
Furthermore, it is multidimensional modeled and is used for the storage of historicized, cleansed, validated, synthesized, operative,
internal and external data. Stakeholders of a data warehouse system are interested in analyzing their business processes in a
comprehensive and flexible way. Mostly they already have a comprehensive understanding of their business processes, which
they want to explore and analyze.
What they actually need is a view of their business processes and its data, which allows them an extensive analysis of their data.
For this purpose data warehouses are modeled multidimensional, which corresponds to a typical view of its users. This analysis
view of the business processes can be very different to the general view even though the underlying process is the same. Hence
it is necessary to elicit requirements from the stakeholder of a data warehouse, which belong to their analysis views. The design
of data warehouse system is highly dependent on these requirements. Very often data warehouses are built without understanding
correctly these needs and requirements and consequently fail for that reason.
During the requirement definition process system analysts of the IT department or consultants work together with stakeholder
and users to describe the requirements for the data warehouse system. The data warehousing team receives these descriptions,
but they have often trouble understanding the business terminology and find the description too informal to use for the
implementation. Therefore the data warehousing team writes its system specification from a technical point of view. When the
system specification is presented to the users, they do not quite understand because it is too technical. They are, however, forced
to accept it in order to move forward.
This approach can easily result in a data warehouse system that does not meet the initially defined requirements because often
the users, the system analysts and developers don’t speak the same language. Such communication problems can make it difficult
Data Management and Decision Support
330 2001 — Seventh Americas Conference on Information Systems
to turn a description of an analysis system into a technical specification of a data warehouse system that all parties can understand.
In addition, because of a technical system specification that is not fully understood by the users, a data warehouse system becomes
too difficult or impractical for the intended purposes. Therefore, it will not deliver the expected effect to the company. In these
cases often departments will develop data marts for their own purposes, which can be considered as stovepipes and makes an
enterprise-wide analysis system impossible.
The challenge is to model a data warehouse system in a way that is both precise and user-friendly. Each symbol describing the
analysis process should be intuitive for the user and have defined semantics, so that the developers can use the description as a
general, but precise specification of the data warehouse system.
In this paper, we use use cases for getting a comprehensive, intuitive specification for the data warehouse system, which is
satisfying and understandable for all involved parties. Use cases have two advantages, which make them suitable for representing
the requirements for a data warehouse system. First, use case diagrams are part of the UML standard and hence are generally
accepted as a notational standard in the software community and second, they can be used on a general level, where
implementation details are completely suppressed.
In the following sections we show a use case driven approach for the development of data warehouse requirements. The remainder
of this paper is organized as follows. In section 2, we discuss the contribution of this paper and related work. In section 3, we
discuss the abstraction levels of data warehouse requirements. In section 4, we argue why to use use cases for capturing data
warehouse requirements. In section 5, we introduced iterations for an incremental and iterative use case development. Finally,
in section 6 we present our conclusion.
Contribution and Related Work
Building a data warehouse is different from developing transaction systems, whereby the requirement analysis process for the
latter is supported by numerous methods (Coad, Yourdon 1991), (Davenport 1993), (Finkelstein 1996), (Partsch 1998). Up to
now the data warehouse design process has not been supported by a formal requirement analysis method although there are some
approaches for requirement gathering.
Inmon (1996) argues that the data warehouse environment is data driven, in comparison to classical systems, which are
requirement driven, and the requirements are understood after it is populated with data and being used by the decision support
analyst. He derives the data model by transferring the corporate data model into a data warehouse schema and by adding
performance factors.
Anahory and Murray (1997) propose a catalogue for conducting user interviews in order to collect end user requirements. They
state that a data warehouse is designed to support the business process rather than specific query requirements, but do not further
discuss their statement.
Another process driven approach is applied by Kimball (1996 and 1998), whereby the fundamental step of the design process is
based on choosing a business process to model. As this approach has proven its success in various projects, and as enterprises
in general have shifted to process-centered organizing, we adopt the process-oriented approach for the basis of our work: a formal
requirement analysis concept for data warehousing.
In this paper we present three different abstraction levels of data warehouse requirements to address better all parties, which are
involved in the requirements development process. In order to enhance the communication between stakeholders, end users, and
the data warehouse development team we focus on modeling the requirements with use cases and derive the detailed system
requirements from them instead of representing the requirements just in specification templates.
Data Warehouse Requirements
Requirements of an enterprise-wide data warehouse system determine its functional behavior and its available information, for
example what data must be accessible, how it is transformed and organized, as well as how it is aggregated or calculated. The
requirements enable the stakeholders to communicate the purpose, establish the direction and set the expectations of information
goals for the enterprise. Stakeholders often express their needs in general expectations of the data warehouse system to improve
their business. This business view describes the goals and expectation of stakeholders, which is the foundation of the data
warehouse requirements. On the other hand, the development team of a data warehouse system expects a complete, correct and
Bruckner et al./Developing Requirements for Data Warehouse Systems with Use Cases
2001 — Seventh Americas Conference on Information Systems 331
Requirement Deliverables
Functional
Requirements
Analysis
Functionality
Data
Staging
Front-End
Requirements
...
Requirement Deliverables
Other
Requirements
Requirements
which are not
included by
functional and
information
requirements
Requirement Deliverables
Information
Requirements
Measures
Analysis
Dimensions
Data Sources
...
Requirement Deliverables
User Requirements
Use
Cases
Test
Cases
Business
Rules
User
Profiles
User
Types
User
Goals
Requirement Deliverables
Other
Attributes
Other
Requirement
Attributes
Requirement Deliverables
Information
Attributes
Information
Quality
Information
Granularity
Information
Security
...
Requirement Deliverables
Functional
Attributes
Performance
Attributes
User Interface
Attributes
Operational
Attributes
...
Security
Attributes
Detailed System
Requirements
Requirement
Attributes
Requirement Deliverables
Business Requirements
Business
Objectives
Vision
Statement
Sope
Definition
Business
Context
System
Context
Success
Factors
Interfaces
Requirements
Environmental
Requirements
Figure 1. Abstraction Levels of Data Warehouse Requirements
unambiguous specification of the
system it has to build, which means a
further refinement of the business
requirements from the stakeholders.
Therefore, it is necessary to transform
the business requirements to a
detailed, testable, and complete
specification for the data warehouse
team (Wiegers 1999).
For that reason, data warehouse
requirements have different
abstraction levels. They include the
following levels shown in Figure 1.
Business Requirements
Business requirements represent
high-level objectives of the
organization for the data warehouse
system. They are captured in a
document describing the project’s
vision and scope. Further services of
the data warehouse system are
derived from the business
requirements, which are represented
in a system context diagram. The
business requirements identify the
primary benefits that the data
warehouse system will provide to the
organization and its users. They
represent the top level of abstraction
in the requirements chain. They
express business opportunities,
business objectives and describe the
typical users and organizations
requirements and their provided value of the system at a high level.
User Requirements
User requirements describe the tasks that the users must be able to accomplish with the help of the data warehouse system. User
requirements must be collected from people who will actually use and work with the data warehouse system. Therefore, these
users can describe both the tasks they need to perform with the data warehouse system and the nonfunctional characteristics that
are important for the data warehouse to be well accepted. The user requirements must align with the context and objectives
established by the business requirements. They are captured in use cases or scenario descriptions; they focus on what the users
need to do with the data warehouse system and are therefore much more powerful than the traditional requirements elicitation
approach of asking users what they want the system to do.
Detailed System Requirements
They represent the data warehouse system requirements on a very detailed level. The high level of detail facilitates the complete,
fine-grained specification of the requirements, which are an important input for the development team. They must align with the
user and business requirements and contain a fit criterion, which allows requirement verification.
Data Management and Decision Support
332 2001 — Seventh Americas Conference on Information Systems
Functional requirements define the functionality that the development team must build into the data warehouse system to enable
users to accomplish their tasks, thereby satisfying the business requirements. Functional requirements capture the intended
behavior of the data warehouse system. This behavior may be expressed as services, tasks or functions the system is required to
perform. They describe what the analysis system must do – an action that the system must take if it is to provide useful
functionality for its users.
Information requirements define the information needs of the company. They describe the information and data, which the data
warehouse should deliver or should have access to. They specify the provided data, by what quality it should have, where it comes
from, how it should be processed, how it should be combined for the analysis process and which analysis methods will be used.
Other requirements besides functional and information requirements can be specified, to describe further relevant aspects of the
data warehouse system, like interface requirements or environmental (cultural, political, legal) requirements.
Requirement Attributes
They augment the description of the functional, information or other requirements by describing the characteristics in various
dimensions that are important either to users or to the data warehouse team. Requirement attributes are properties, or qualities
that the data warehouse system should have. These might include standards, regulations, and conditions to which the data
warehouse system must conform; description of external interfaces; performance requirements, design and implementation
constraints; and quality attributes. Requirement attributes are usually attached to the detailed system service requirements. For
instance, when the functional requirements are known, it can be determined how they are to behave, what qualities they are to
have, and how big or fast they should be. When the information requirements are known, its attributes can be determined, like
data quality or data granularity.
Functional vs. Information Requirements
Traditional software requirements distinguish between two types of requirements: 1) functional requirements and 2) non-
functional requirements. This separation of requirements can be applied also to data warehouse systems. But the heart of data
warehouse system is the data and information that is delivered. Information requirements are included in traditional software
requirements in the functional and non-functional requirements.
Because data warehouses are built for delivering data, the requirements, which describe this data, are highly significant for the
data warehouse design and more comprehensive than the requirements in traditional software systems. Therefore, traditional
software requirements types are not very suitable for data warehouse projects. In order to address the aspect that data warehouse
systems are information-centric, we explicitly distinguish between functional requirements and information requirements.
The separation of functional and information requirements can be very advantageous for the data modeler of a data warehouse.
Because primarily only information requirements are relevant for the data model of a data warehouse, the data modeler isn’t
diverted by other requirements.
Requirements Gathering with Use Cases
For many years, system analysts have used scenarios to describe ways a user can interact with a software system to help elicit
requirements. Ivar Jacobson (1992 and 1995) and others formalized this into the use case approach to requirements elicitation and
modeling. Although use cases emerged from the object-oriented development world, they can be applied to projects that follow
any development approach, because the user does not care how to build the system. Therefore, the shift in perspective and thought
processes that use cases bring to requirements development is more important than drawing formal use case diagrams. The focus
on what the users need to do with the system is much more powerful than other traditional elicitation approaches of asking users
what they want the system to do.
Use Cases and Detailed System Requirements
The use case descriptions often do not provide the data warehouse team with enough detail about the functionality they must build
and about the information that must be made available. Gathering all requirements in use case descriptions can result in
cumbersome and complex use cases. On the other hand, if stopping requirements development at the user requirements stage, at
Bruckner et al./Developing Requirements for Data Warehouse Systems with Use Cases
2001 — Seventh Americas Conference on Information Systems 333
construction time the data warehouse team will have many questions to fill their information gaps. To reduce this uncertainty,
each use case needs to be elaborated into its detailed system requirements (Arlow 1998). For very small use cases, the use case
description can be sufficient and the elaboration of the detailed system requirements can be skipped.
Each use case leads to a number of detailed system requirements that will enable a data warehouse user to perform their pertinent
task, whereby several use cases might need the same detailed system requirement. For example, if three use cases require the same
data with the same quality, the specification of the data should not be written down in every use case. The specification of the
data and its quality can be described by detailed system requirements, which can be associated with a use case in several ways.
The approach that is taken depends on whether the data warehouse team should perform the design, construction and testing from
use case documents, from detailed system requirements, or from a combination of both. None of these methods is perfect and it
has its advantages and disadvantages. By selecting an appropriate approach, it is important to avoid any duplicated information
in multiple locations, as redundancy makes requirements management more difficult.
Use Case Refinement
Use case refinement activities ensure that requirements are accurate, complete, and demonstrate the desired quality characteristics.
Data warehouse requirements that seem fine when they are captured by use cases might turn out to have problems when the data
warehouse team works with them. If for instance test cases for use cases are defined, possible ambiguities and vagueness in some
of the use cases may be found. They must be removed if the data warehouse requirements are to serve as a reliable foundation
for the design and implementation. These use case refinement activities include activities to ensure that:
— The use cases describe the intended system behavior and characteristics of the data warehouse system.
— The use cases were correctly derived from the business requirements or other origins.
— The use cases are complete and of high quality.
— All views of the use cases are consistent.
— The use cases provide an adequate basis to proceed with the design and implementation of the data warehouse system.
Iterative and Incremental Use Case Development
For the development of use cases, we suggest an iterative and incremental approach. An incremental approach to requirements
specification includes completing batches of use cases together, but not assuming that all artifacts must be completed at the same
time. An iterative approach to requirements specification includes the refinement of use cases through a number of iterations.
Iterative steps in gathering requirements with use cases are highly dependent on individual situations. We can’t make a suggestion
about the exact number of iterations that are needed to complete the requirements in all situations. However, it is possible to say
that iterative requirements specification always proceeds through the same logical steps in every situation. Kulak and Guiney
(2000) introduced an approach for an iterative use case development. We extended this approach to be more suitable for data
warehouse environments. We suggest following four logical steps for the use case development:
Table 1. Use Case Iterations
Iteration Description
Facade Outline and high-level descriptions
Filled Broadening and deepening
Analyzed Evaluation, narrowing and pruning
Optimized Reconciliation, reassessment, enhancement
Finished Touching up and fine-tuning
Facade Iteration
The Facade Iteration is the first iteration in the requirements lifecycle of use cases. Its purpose is to create placeholders for each
major interaction of a user with the data warehouse system. A Facade use case contains only the minimum information as a
placeholder is gathered, which includes names and a short description of each user interaction with the data warehouse system.
Data Management and Decision Support
334 2001 — Seventh Americas Conference on Information Systems
It also identifies the major actors of the system. Defining use cases for this iteration is difficult, because you don’t have any
concept of the data warehouse system yet. For this reason, the data warehouse team and the stakeholders will do their best work
if the environment encourages openness and creativity.
Filled Iteration
The objective of the Filled Iteration is to sharpen the ideas of the use cases from the Facade Iteration and to create a
comprehensive set of use cases and business rules that describe the data warehouse system. As you delve into more detail during
this iteration, it is possible to find Facade use cases which are too general and need to be broken down into several Filled use
cases. During this iteration, most of the time is spent to understand the requirements from the stakeholders, and exploring possible
solutions for the data warehouse system. Although these requirements are rough and unpolished, they contain details and include
functional and information requirements as well requirement attributes.
Analyzed Iteration
The requirements from the previous iterations can be discouragingly large. Therefore, in the Analyzed Iteration only the best
options are selected. The Analyzed Iteration clears a path through the existing requirements and leaves clear project requirements.
It separates the essential from the nice-to-have. At the end of this iteration, there is sufficient information and material to build
a successful data warehouse system.
Use case analysis involves refining, evaluating, and scrutinizing the gathered requirements to make sure all stakeholders
understand what they mean and to find error, omissions, or other deficiencies. The goal is to develop use cases of sufficient quality
and detail, so that they can be used to construct realistic project estimates and to proceed with the design, construction, and testing
of the data warehouse system. Each use case is reviewed to make sure that the descriptions are complete and provide sufficient
information without being too detailed or vague. This will be achieved by continual formalizing the use cases by specifying
priorities, traceability links, fit criteria, pre- and postconditions, triggers, and inputs/outputs of the use case etc. Furthermore, in
this iteration use cases are inspected, whether there are possible improvements or optimizations of the event course.
Optimized Iteration
The objective of the Optimized Iteration is an inspection for conflicting, inconsistent or missing use cases. The difference between
this iteration and the previous iteration is, that in the Analyzed Iteration, each use case is checked independently, whereas the this
iteration considers all of the use cases and requirements and their effect on each other. The input of this iteration are all use cases,
whereas the input of the Analyzed Iteration are the individual use cases.
A further goal of this iteration is a new assessment of the complete specification of the data warehouse system:
— Remeasurement of the effort required building the data warehouse system. The initially estimated budget, schedule, and staff
for building the data warehouse system, can be adjusted according to the final requirements specification.
— Reanalyzing the risks involved in completing the data warehouse project. Due to the earlier requirements elicitation and
analysis activities and thus a better understanding of the functionality and impacts of the data warehouse system for the
business, the risks for the data warehouse project can be assessed more accurately.
— Reassessment of the go/no-go criteria, which where defined in the project blast-off
Finished Iteration
In the Finished Iteration the stakeholders and the requirements analysts decide if the use cases are correct and complete to go on
with the design or implementation of the data warehouse system. The objective of this iteration is to make reasonable decisions
to craft requirement attributes, which are important for the data warehouse design and implementation. Finished use cases are the
primary input to the design phase of the data warehouse system. They influence the design of a data warehouse system by
communicating the scope and comprehensive descriptions of what the users want from the data warehouse system.
Bruckner et al./Developing Requirements for Data Warehouse Systems with Use Cases
2001 — Seventh Americas Conference on Information Systems 335
Conclusion
The use case model is an excellent means of both expressing user analysis activities and providing a comprehensive picture of
what the data warehouse is intended to perform. In this paper we presented a use case driven approach for the development of
data warehouse requirements. We specified data warehouse requirements by different abstraction levels, which allow a
straightforward development of the requirement specification. Because data warehouses are information-centric, we stated, that
there is a need to separate information requirements as individual requirement type. We discussed the iterations for a use case
development, which facilitate the modelling of user requirements for data warehouse systems.
References
Anahory, S. Murray, D., Data Warehousing in the real World – A practical Guide for Building Decision Support Systems,
Addison-Wesley Publishing 1997.
Arlow, Jim, Use Cases, UML Visual Modeling and the Trivialisation of Business Requirements, Requirements Engineering
Coad, P., Yourdon, E., Object-Oriented Analysis. Englewood Cliffs, N.J., Prentice Hall 1991.
Davenport, Th., Process Innovation: Reengineering Work through Information Technology, Harvard Business School Press 1993.
Finkelstein, A., Requirements Engineering Research: Coordination and Infrastructure , Requirements Engineering 1, 1996.
Inmon, W. H., Building the Data Warehouse, Wiley & Sons 1996.
Jacobson, I., Christerson, M., Jonsson, P., Overgaard, G., Object-Oriented – Software Engineering – A Use Case Driven
Approach, Addison Wesley 1992.
Jacobson, I., Ericson, M., Jacobson, A., The Object Advantage – Business Process Reengineering with Object Technology, ACM
Press, Addison-Wesley Publishing 1995.
Kimball, R., The Data Warehouse Toolkit: Practical Techniques For Building Dimensional Data Warehouse, John Wiley & Sons
1996.
Kimball, R., Reeves, L., Ross, M., Thornthwaite, W., The Data Warehouse Lifecycle Toolkit, John Wiley & Sons, 1998.
Kulak, D.; Guiney E., Use Cases – Requirements in Context, Addison Wesley, New York, 2000.
Partsch, H., Requirements Engineering (in German), Springer 1998.
Wiegers, Karl, E., Software Requirements, Microsoft Press, 1999.
