Challenges in Requirements Engineering:

1
Challenges in Requirements Engineering A
Research Agenda for Conceptual Modeling Salvat
ore T. March (Presenter) Vanderbilt
University Gove N. Allen Brigham Young University
2
Design
Representation
Specification
Abstract
Implement
IS
3
Business Domains Include
People Employees Products Financial
Obligations Corporations Customers Contracts Sales
Orders Promotions Discounts Money Machines
Trucks Purchase Orders Government
Agencies Vendors Purchase Orders Complaints Organi
zational Structure Production Schedule Raw
Materials Buildings Job Descriptions Skills
4
Conceptual Modeling Grammar
Theoretical Basis Ontology (Bunge, Davidson,
Searle) Epistemology Concept Theory Speech Act
Theory Semiotics Empirical Basis Ease of
Use Understandability Analyst Performance Domain
Expert Performance Utility in Communication of
Meaning
5
Bunge's Ontology

• The world is composed of concrete objects
  (things) that possess substantial properties.
• All and only things possess the property of
  "existing really."
• The state of a thing is the totality of its
  properties.
• The set of things is closed.
• All things possess the universal property "energy
  which measures changeability.
• An event is the change in state of a thing.
• Change is governed by immutable natural laws.
• A kind is the set of things each of which
  possesses a common set of properties.

6
Searle's Ontology

• The world is comprised of concrete (physical) and
  conceptual (socially constructed) objects.
• Socially constructed objects are created by human
  collective agreement. They exist "in human minds"
  and serve human purposes.
• Brute facts describe properties possessed by
  concrete objects. Institutional facts ascribe
  deontic powers (rights, obligations, duties) that
  form the basis for social intercourse (action).
• Changes in institutional facts are governed by
  human created and mutable rules.
• Institutional facts are ontologically subjective
  but epistemologically objective.

7
Davidson's Ontology

• Identification of objects
• we ought not to countenance entities unless we
  are prepared to make sense of sentences affirming
  and denying identity of such entities." (1980, p.
  164)
• Events are individuated objects
• I suggested that sentences about events and
  actions be construed as requiring an ontology of
  particular, unrepeatable, dated events.
• Understanding human action and communication
  requires the recognition of events as objects
• I do not believe we can give a cogent account of
  action, of explanation, of causality, or of the
  relation between the mental and the physical,
  unless we accept events as individuals.

8
Ontological Objects

• Concrete Objects (Bunge)
• Exist in space and time
• Possess observer-independent properties
• Governed by natural laws
• Composed and decomposed not created or destroyed

• Conceptual Objects (Searle)
• Exist by "collective acceptance"
• Are ascribed observer-dependent attributes
• Governed by human rules
• Created and destroyed by human endeavor

9
Systems Taxonomy
Material Systems - consist of concrete objects
that can be perceived (exist in space and time).
They possess scrutable properties represented by
substantial attributes (e.g., height, weight,
density). The state of a concrete object is the
totality of its properties at a point in time.
Properties change in accordance with discovered
natural laws. Social Systems - consist of
concrete and conceptual objects (exist by
collective agreement). Attributes are invented
and ascribed to concrete and conceptual objects
to enable social intercourse (e.g., name, job
title, price, revenue). The status of an object
is the set of values of its attributes at a point
in time. Invented attributes change according to
invented rules initiated when events occurs.
10
Research Agenda

• An Ontology of the Artificial
• Build
• Constructs
• Models
• Methods
• Instantiations
• Evaluate
• Theoretical
• Empirical

11
Premises

• Conceptual modeling is primarily concerned with
  representing social phenomena.
• Defined social phenomena, termed conceptual
  events, are governed by invented rules.
• An information system is primarily an
  event-processing mechanism.
• The scope of a conceptual event rule is the set
  of social institutions with the authority to
  change and enforce it.
• The range of a conceptual event rule is the set
  of objects that are affected by it.
• Identification of objects (conceptual and
  concrete) is fundamental to conceptual modeling.

12
Active Information Systems
Application of Rules
Application of Rules
States Events
IS
Record
Events Status Rules
State History Interpretation
Report
Event-Processing Mechanism
13
Ontology of the Artificial

• Objects Concrete and Conceptual Simple and
  Composite
• Attributes Substantial and Invented
• Object Types (Classification) Natural
  (Property-bases) and Artificial (Attribute-based)
• Status Substantial and Invented Attributes
• Events Concrete and Conceptual Causal and
  Relational Simple and Composite Immutable
  Natural Laws and Mutable Invented Rules

14
Research Agenda

• Representation and Classification of Events and
  Rules
• Generation and Evaluation of Rules that Draw
  Inferences from Events
• Conceptual Modeling of Possible Futures Based on
  Proposed or Conjectured Events
• Conceptual Modeling of Temporal Dependencies
  Among Events