Browse Prior Art Database

Intelligent Object-Oriented Scenario Generation

IP.com Disclosure Number: IPCOM000117046D
Original Publication Date: 1995-Dec-01
Included in the Prior Art Database: 2005-Mar-31
Document File: 6 page(s) / 177K

Publishing Venue

IBM

Related People

Lightstone, SS: AUTHOR

Abstract

Described is a technique for intelligently generating scenarios. Automated scenario generation is useful for a number of computing tasks, including soft ware testing and computer simulation. The technique is based on a combination of three well known concepts: o Finite State Machine (FSM) Modelling, o Object Oriented Design, and o Probability modelling for operations.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 44% of the total text.

Intelligent Object-Oriented Scenario Generation

      Described is a technique for intelligently generating
scenarios.  Automated scenario generation is useful for a number of
computing tasks, including soft ware testing and computer simulation.
The technique is based on a combination of three well known concepts:
  o  Finite State Machine (FSM) Modelling,
  o  Object Oriented Design, and
  o  Probability modelling for operations.

      Given a set of operations that can be performed on a system,
and a set of object types that can exist within the system, a Finite
State Machine can be built to allow a computer to intelligently
select applicable operations against states in the system.  The FSM
is modelled after the potential combinations of objects that may
exist in the system.  Each possible combination of objects represents
a new state in the FSM.  The intelligence of this scenario generation
is enhanced by applying a probability distribution against the
selection criteria for these operations.  Thus, the computer not only
knows what operation are allowable, but which are most likely to
occur.

      Once all of the models have been defined, a simple template of
rules is developed from which all of the information required to
intelligently generate scenarios is derived.

      The first task is the development of the FSM.  Using object
oriented design methodologies, a FSM is designed based on the
presence or absence of objects.  Fig. 1 graphically represents such a
FSM.  In this sample FM, each state in the finite state machine is
assigned a state number which is a bitmap of existing objects in that
state.

      The next task lists the operations that can be performed on the
system at any state, an d the assignment of probabilities to these
operations.  Fig. 2 shows a sample distribution of operations for an
imaginary image processing product.

      In normal processing some operations will create objects, other
operations may destroy objects, while some operations will not affect
the existence of objects at all.  The creation or destruction of
objects has the effect of changing the state of the system in the
FSM.

      With both the FSM model and the list of operations complete,
each operation in the list is compared with the FSM to determine what
objects must exist for each operation in the list to be executable.
This minimum required set of objects is called the FSM Requirement
for the operation.  Each operation must have a FSM Requirement
associated with it.  Operations may share equal FSM Requirements.  An
FSM Requirement need not be a state that exists in the FSM.

      For an operation to be performed, it must have a FSM
Requirement that is a reachable state in the FSM, or a substate of a
reachable FSM state.  In this context, a substate is defined as being
any state for which all of its composite objects exist in a state
within the FSM.  For example, consider a state where t...