FROM STRATEGIC TO CONCEPTUAL ENTERPRISE
INFORMATION REQUIREMENTS
A Mapping Tool
Gianmario Motta and Giovanni Pignatelli
Information and Systems Department, University of Pavia, Via Ferrata n.1, I-27100 Pavia, Italy
Keywords: Strategic Information Requirements, Aggregated Business Entities, Systems Requirements Engineering,
Requirements Analysis, Systems Analysis, IT Strategic Planning, IT Strategy.
Abstract: Enterprise Information analysis can be modeled on three levels: Logical, Conceptual and Strategic. Logical
level is used daily on thousand of projects to design databases. Conceptual level is used by analysts to
structure detailed information needs expressed by users. Strategic level is used by IT and user management
to define Enterprise Information Architecture and/or to assess the viability of the current information assets.
While mapping conceptual onto logical modeling is well-established, strategic and conceptual levels are
poorly linked. This drawback very often prevents enterprise to implement a sound information strategy. We
here present a method that maps strategic enterprise information into conceptual information modeling. For
strategic modeling a comprehensive framework is used that enables to readily identify information domains
of a wide range of enterprises. Mapping strategic to conceptual models is performed by a set of simple and
predefined rules. The paper also illustrates the tool that has been developed to assist the whole design and
mapping process. Finally a case study on materials handling exemplifies our approach.
1 INFORMATION
REQUIREMENTS LEVELS
It is a common practice to classify enterprise
information requirements in two abstraction levels,
logical and conceptual. The former is represented by
Relational models (Elmasri, 2004) and the latter by
Entity Relationship (ER). Of course, over years
other modeling techniques have been developed. For
instance Dimensional Fact Model (Golfarelli, 1998)
focuses at the conceptual level. Other modeling
technique families as Unified Modeling Language
(Object Management Group, 2005) cross levels.
Each abstraction level is for a certain community
of users. Typically the conceptual level is for
analysts and logical one is for Database
Administrators (DBA) or generally for
implementation engineers. However, ER is too
detailed for an overview of an even limited domain.
Actually an ER schema of the customer/order
domain in an enterprise may count hundreds of
entities! Therefore, above the conceptual level, a
third abstraction level is needed. This third level
shall address enterprise information as a whole and
also target the strategic needs that are of key interest
for management. In short it shall be the backbone of
IT strategy and planning. The positions of the three
abstraction levels are represented by Figure 1.
Figure 1: Abstraction levels of enterprise Information.
In short, the strategic level should represent
information in a very aggregated and compact form
that can be understood by IT and user management
and can be used in IT strategic planning. The
strategic level issue is in Enterprise Information
Architecture (
Josey, 2009) and in Enterprise
Information Integration (EII). The latter has the
440
Motta G. and Pignatelli G. (2010).
FROM STRATEGIC TO CONCEPTUAL ENTERPRISE INFORMATION REQUIREMENTS - A Mapping Tool.
In Proceedings of the 12th International Conference on Enterprise Information Systems - Information Systems Analysis and Specification, pages
440-446
DOI: 10.5220/0002893804400446
Copyright
c
 SciTePress

purpose of combining into an unified format
information from diverse sources (Bernstein 2008;
Halevy, 2005).
Strategic Information requirements have been
specifically addressed by Enhanced Telecom
Operations Map® (eTOM). The eTOM framework
(TMForum, 2003) includes the Shared Information
Data model (SID) that offers a normative paradigm
for shared information / data, based on the concepts
of Business Entities and Attributes (TMForum 2003,
2005). A Business Entity is a thing of interest to the
business, while Attributes are facts that describe the
entity. In short “an Aggregate Business Entity
(ABE) is a well-defined set of information and
operations that characterize a highly cohesive,
loosely coupled set of business entities”. The
framework is shown in Figure 2 where each block
represents an Aggregated Business Entity. By
defining ABEs in telecommunications domain, SID
is a normative framework for information but it
lacks universality, since it is oriented to
telecommunications nor it provides an axiomatic
approach to identify Entities.
Figure 2: SID, TMForum 2005.
Actually a really universal, not domain specific,
approach to Strategic Information Requirements
modeling is an issue since the heydays of
information systems and it has been addressed by
different techniques families.
The analyst uses these methods to structure
information needs, gathered from management,
through interviews or other sources. Business
Systems Planning (BSP), very popular in Eighties
(IBM 1975) associates data classes and processes in
a grid, that shows which process uses which data.
Later, Information Strategy Planning - ISP (Martin
1990) integrates different information models, such
as BSP, Entity Relationships and Data Flow
Diagrams (DFD). Finally, with the success of ERP
(Enterprise Resource Planning) software, a new
family of information analysis techniques integrates
information, processes and organizational structures,
such as the successful ARIS (Architecture of
Integrated Information Systems), that provides some
normative definition of high level information, but it
mirrors SAP (Scheer, 2000).
With this same purpose of universality, a recent
technique called Strategic Information Requirements
Elicitation - SIRE (Motta, 2008) contains a universal
catalogue of enterprise Strategic Information
Entities-SIE (Table 1). Each SIE results from
crossing Information Types and Information
Domains. Information Types reflect the nature of
Information that may be structural (Master Data) or
describe events (Transaction Data) or define
computed indicators (Analysis Data). In turn,
Information Domains describe the universe about
which information recorded and are conceptually
similar to SID’s ABEs.
By a sequence of steps, Strategic Information
Entities (SIE) are tailored to a specific enterprise.
Potentially, SIRE offers a flexible approach that can
be incorporated in methodological containers as
TOGAF. Actually it is not bounded to a specific
industry and customization method comply specific
needs.
Table 1: SIRE catalogue of Strategic Information Entities.
INFORMATION TYPE
Master
Data
Transaction
Data
Analysis
Data
INFORMSTION DOMAIN
Stakeholders
Law
Competitor
Customer
Supplier
Broker
Shareholder
Resources
Personnel
Plants
Raw
materials
Cash
Context
Structure
Project
Region
Output
Process
Product
Service
FROM STRATEGIC TO CONCEPTUAL ENTERPRISE INFORMATION REQUIREMENTS - A Mapping Tool
441



Figure 3: Stages of the Strategic-to-conceptual mapping.
Table 2: Mapping steps.
Step Input Output Activities
Selection
Standard
Information
Catalogue
Selected
Strategic
Information
Entities
a) Define the scope of
analysis
b) Select SIEs and add
properties
Customization
and
Refinement
Selected
Strategic
Information
Entities
Customized
Strategic
Information
Entities
Creation /
Specialization /
Decomposition of
Strategic entities
Course
Mapping
Customized
Strategic
Information
Entities
Conceptual
Information
Islands
Link Strategic Master
Data to Strategic
Transaction Data
Link of
Information
Islands
Conceptual
Information
Islands
Conceptual
Linked Entities
Link between
Information Domain
Refinement
Conceptual
Linked
Entities
Refined
Conceptual
Entities
a) Creation /
Specialization /
Decomposition of
Conceptual entities
b) Creation of new
Relationship as needed
Our work intends to illustrate a model and a tool
to map SIE at strategic level down to the Entities at
conceptual level. In an ideal Information
Engineering, each abstraction level can be mapped
over the lower one. Therefore in our level pyramid
(Figure 1) we should have two top-down mappings
namely, the strategic-to-conceptual and the
conceptual-to-logical. Numberless model-to-model
transformation techniques and tools have been
developed and used for conceptual-to-logical
mapping. Our purpose is to present a technique and
tool to support strategic-to-conceptual mapping.
2 THE MAPPING METHOD
This section addresses the transformation of the
Strategic Information Catalogue in a standard ER
Schema. In order to obtain a viable conceptual
schema from the initial catalogue of SIEs we have
defined six abstraction stages (Figure 3). The first
three stages actually fall into the area of strategic
information requirements elicitation and have been
already illustrated (Motta, 2008). The subsequent
stages fall within conceptual analysis and are based
on ER schema. Table 2 summarizes input, output
and activities of each step (Motta, 2009).
The model transformation includes three steps
(1) Course Mapping, (2) Link of Information
Islands, (3) Refinement. Mapping has in input the
SIE model and converts it into a preliminary ER
model. The mapping is applied for each domain in
the model and could be summarized in Table 3.
Through this step analysts delete the “horizontal
discontinuities”, and link master and transaction data
within the same information domain.
The first step is domain-centered and the
preliminary ER schema is made of low-coupled
“Conceptual Information Islands”. An Information
Islands is the association between a Master Strategic
Entity and its related Transaction Strategic Entities.
The name island underlies that no cross domain
links exist but only master-to-transaction.
ICEIS 2010 - 12th International Conference on Enterprise Information Systems
442

Figure 4: SIE meta-model (UML class notation).
As a second step the analyst links Information
Islands by identifying common entities, attributes
and relations between domains and obtains a more
cohesive ER schema. By this step the analyst
overcomes “vertical discontinuities” between master
information belonging to different information
domains and also “diagonal discontinuities” between
master information and transaction information
belonging to different information domains. As a
third step the analyst inserts new relations between
entities, specializes or decomposes entities and
attributes and aggregates or generalizes entities. The
last step enhances the ER schema by inserting
deeper domain competences.
Table 3: Mapping algorithm of SIE to ER model.
SIRE Model ER Model
Specialization
Enhanced ER Specialization (Overlap
or Disjoint)
Decomposition
Compound ER or Compound/Complex
attribute
Property of Master Data Entity Type or Attributes
Property of Transaction
Data
Entity Type and Relationship Type
Property of Analysis Data Calculated attributes
3 TOOL
In order to support the analysts in identifying
Strategic Information requirements and map them on
the Conceptual level, we have designed an Eclipse
based tool that enables the creation of well-formed
SIRE models and related ER schemas. The tool
follows the six steps discussed above. The
development of such a tool overcomes the scarcity
of tools for modeling conceptual information level
and moreover for mapping strategic level into
conceptual level.
Data Tool Platform (DTP) project
(http://www.eclipse.org/datatools/) is a powerful
Eclipse project but it produces only relational
schemas. Eclipse plug-in central (http://
www.eclipseplugincentral.com) provides 28 plug-ins
for relational modelling such as CLAY MARK II,
ERMaste, AMATERAS ERD and Mogwai
ERDesigner. Alas no conceptual modelling tool is
provided. The commercial DATABASE VISUAL
ARCHITECT (http://www.visual-paradigm.com/)
provides the design conceptual models but it is not
open and it does not map strategic requirements on
conceptual level.
Our tool is based on the SIE meta-model that has
been developed with the Graphical Modeling
Framework (GMF, http://www.eclipse.org/gmf/)
provided by the Eclipse platform (Figure 4). The
upper entity classes represent the catalogue
containers and are ancillary classes. In the lower
row are represented the super-types respectively the
[Information] Domain, [Information] Type and SIE
entity.
The conceptual meta-model (Figure 5) reflects
the well-known Elmasri’s (Elmasri, 2004)
representation and, therefore does not need to be
explained in detail.
FROM STRATEGIC TO CONCEPTUAL ENTERPRISE INFORMATION REQUIREMENTS - A Mapping Tool
443

Figure 5: ER metamodel (UML class notation).
3.1 The Case Study
To illustrate how the mapping method and its related
tool work we will use a very easy case study.
MACINE manufactures about 50,000 tractors.
Products models are about 2.000, and result from
combinations of base models and optional parts.
Production is performed in one plant, and products
are sold by dealers. Below we describe materials
handling operations; candidate entities are in capital
letters.
3.1.1 Receiving
Suppliers ship pallets according the deadlines
specified on MONTHLY SUPPLY PLAN (MSP).
Shipments are described a BILL OF MATERIAL
(BOM). The formal correctness of supplier’s
information is checked against information stored in
PART, supplier orders (ORDER) and SUPPLIER.
An ENTRY BILL (ENB) for is issued, in which are
recorded the details of the arrival. For each received
pallet, a LOADING UNIT record is created. It
specifies details of delivery. A paper copy of
LOADING UNIT (LUN) follows the material.
Quantity and quality differences are recorded on
the ENTRY BILL, ORDER and PART files.
Differences in quality are recorded on a DISCARD
BILL (DSB) and related material is put in an ad hoc
area.
Hot requests (MATERIAL REQUEST) are
flashing in the arrival area and are satisfied by
immediately dispatching the arrived material to the
plant, with a paper copy of MATERIAL REQUEST
(MRQ). Direct dispatch is documented by the same
record as for picking, described below.
3.1.2 Storage
The material is moved to the warehouse entry. Free
warehouse cells are identified and reserved on the
warehouse map (MAP). The material is stocked in
the warehouse and its actual position is recorded on
LOADING UNIT and MAP.
3.1.3 Picking
The production requires picking by a MATERIAL
REQUEST form. The location is identified on
PART and MAP. If the material is found, the
corresponding LOADING UNIT are booked. If the
material is not found, the MATERIAL REQUEST is
forwarded to the receiving staff. When picking is
over, related LOADING UNIT are "erased"; also
MAP (locations vacated) and PART are updated.
Figure 6 reflects the third step of the method in
which the analyst is customizing the Strategic
information entities of the case study. At this point
the analyst can launch the transformation of his table
of customized strategic information entities in
course conceptual entities. Transformation
associates to each Master information its related
Transaction. Specifically you have three Master
entities that, respectively, reflect the domains of
Resources-Material, Stakeholder-Supplier and
ICEIS 2010 - 12th International Conference on Enterprise Information Systems
444

Figure 6: A screenshot of the tool representing the customized strategic entities.
Figure 7: A screenshot of the linked conceptual entities.
FROM STRATEGIC TO CONCEPTUAL ENTERPRISE INFORMATION REQUIREMENTS - A Mapping Tool
445

Resources-Plant. Each Master entity is related in 1-
to-N associations to a set of Transaction entities. The
subsequent model of linked entities is obtained
through a manual linkage (Figure 8). The links
typically reflect a multiple association of a
transaction entity to master entities. In our case
study this happens for BOM and ORDER (both
linked to PART and SUPPLIER) and for LUN
(linked to PART and MAP. Though links are
identified by domain knowledge, we are planning to
develop a guidance that helps the analyst in
identifying potential cross-domain links already at
strategic level and implementing them at conceptual
level.
Finally the analyst will refine the schema by
well-known ER elementary operations (e.g.
specialization, N-ary relationships, etc…).
4 CONCLUSIONS
We have illustrated a technique and a tool that
enables to design a complete and consistent
Enterprise Information Architecture from the
strategic level down to the conceptual level. The
approach is consistent as is based on robust models.
Actually strategic modelling is based on a normative
framework (SIRE, Motta, 2008) that generalizes
some concepts extensively tested by eTOM SID
(TMForum, 2005). In turn the conceptual level uses
the universally known ER notation. The approach is
complete as it provides simple rules to map the
strategic level onto the conceptual level.
Furthermore the approach can be usefully
integrated in more general frameworks such as
TOGAF and it is supported by a open source tool
based on the Eclipse suite.
Future developments include extended coverage,
tool enhancement and extended validation.
The coverage will be extended by the
introduction of a bottom-up mapping from
conceptual to strategic level. This mapping could
help IT management to extract a strategic view from
the current heterogeneous and diverse databases. A
very similar research direction is to structure a
strategic information architecture from unstructured
text documents (e.g. manuals, organization charts,
interviews and alike).
The tool enhancement will include the
integration of the conceptual modelling tool with
logical modelling tools and, also, guidance in model-
to-model transformation.
Finally we are planning to apply out approach to
other real-life cases mainly in service and
government sectors.
REFERENCES
Bernstein, P. A. and Haas, L. M. 2008. “Information
integration in the enterprise. Communications of ACM
51, 9 (Sep. 2008), 72-79. DOI = http://doi.acm.org/
10.1145/1378727.1378745
Colter, M. A. 1984, “A comparative analysis of systems
analysis techniques”, MIS Quarterly, Vol. 8, N.1,
March
Elmasri, Navathe, 2004, Fundamentals of Database
Systems, Fourth edition, Pearson Education
Matteo Golfarelli , Dario Maio , Stefano Rizzi, Conceptual
Design of Data Warehouses from E/R Schema,
Proceedings of the Thirty-First Annual Hawaii
International Conference on System Sciences-Volume
7, p.334, January 06-09, 1998
Halevy, A. Y., Ashish, N., Bitton, D., Carey, M., Draper,
D., Pollock, J., Rosenthal, A., and Sikka, V. 2005.
“Enterprise information integration: successes,
challenges and controversies” Proceedings of the 2005
ACM SIGMOD International Conference on
Management of Data (Baltimore, Maryland, June 14 -
16, 2005). SIGMOD '05. ACM, New York, NY, 778-
787. DOI= http://doi.acm.org/10.1145/
1066157.1066246
IBM, 1975, Business Systems Planning, GE 20-0257-1
Josey, A., Harrison R., 2009, TOGAF Version 9: A Pocket
Guide, Van Haren Publishing
Martin J., 1990, Information engineering, Prentice Hall,
New York
Motta G, Pignatelli G 2008. Strategic Modelling of
Enterprise Information Requirements. Proceedings of
the 10th International Conference on Enterprise Infor-
mation Systems (Barcelona, Spain, June 12 -16, 2008)
Motta G., Pignatelli G., Information strategy: an analysis
framework, Journal of US-China Public
Administration, 2009
Object Management Group, Unified Modelling Language.
Version 2.0., http://www.omg.org/docs/formal/05-07-
04.pdf
Scheer, W. 2000 ARIS – Business Process Modelling, 3d
edition, Springer, Berlin
TMForum, 2003, Shared Information/Data(SID) Model -
Concepts, Principles, and Domains- GB922, July 2003
TMForum, 2005, Enhanced Telecom Operations Map
(eTOM), The Business Process Framework – GB921,
November
ICEIS 2010 - 12th International Conference on Enterprise Information Systems
446