A COMPUTER SIMULATION MODEL OF LIBRARY OPERATIONS
| Published date | 01 January 1975 |
| Pages | 1-18 |
| DOI | https://doi.org/10.1108/eb026589 |
| Date | 01 January 1975 |
| Author | PAULINE A. THOMAS,STEPHEN E. ROBERTSON |
| Subject Matter | Information & knowledge management,Library & information science |
THE
Journal of
Documentation
VOLUME 31 NUMBER 1 MARCH 1975
A COMPUTER SIMULATION MODEL
OF LIBRARY OPERATIONS
PAULINE A. THOMAS
Aslib
Research and Development Department
and STEPHEN E. ROBERTSON
School
of
Library,
Archive and Information
Studies,
University College London
INTRODUCTION
Aslib is interested in developing modular simulation models of aspects of
library systems in order to investigate the
use
of such models
as aids
to systems
design and operation. The conceptual model of library procedures, deve-
loped during the OSTI-supportcd project on the use of bibliographic
records in libraries, offered a suitable basis for a pilot study. The pilot study
was aimed at investigating the practicality of developing and using simula-
tion models of library operations, and gaining experience in the application
of simulation techniques to problems in the library and information field.
The simulation language used was
GPSS,
in accordance with the
1968
edition
of the IBM GPSS/360 User's Manual, H20-0326-2.
COMPUTER SIMULATION is an Operational Research technique that uses a
model to study the performance or behaviour of
a
system over
time.
We simulate
systems because we want to understand how they work, and to predict how they
might react to change. A detailed description of the behaviour of a system is
translated into
a
computer simulation
model,
and
this is
manipulated to reproduce
the operation of the system over the required period of simulated time. The
likely effects of changes to the system can be predicted by using the model and
only introduced into the real world system if the predicted results are satisfactory.
The model can therefore be used both to supply information on the performance
of an existing system, and as a predictive device to test alternative solutions to
operational problems.
Advantages inherent in this approach include those derived from the use of
models generally:
the ability to study the effects of operations over
time,
and to experiment with
changes in model design without disturbing the actual system;
1
JOURNAL OF DOCUMENTATION Vol.
31,
no. 1
the opportunity to repeat the situation and try new tactics if the first were
unsuccessful;
the projection of the operations of a new system;
and the provision of artificial experience in a training situation.
Studies of organizational behaviour, with their need to accommodate the rich
complexity of interactions between
humans,
and between structural and machine
components and humans, are particularly amenable to interdisciplinary model
building approaches.1 Verbal theories tend to give place to more quantitative,
precise explanations as disciplines develop. At the present stage of organizational
analysis, there
is
scope, and perhaps an urgent need, for modelling with combina-
tions of verbal, graphical, and mathematical elements. Situations too complex to
be solved easily by purely mathematical techniques, can be described and
modelled in a computer simulation, while relevant mathematical functions may
be included as required. The resulting greater intelligibility to the non-mathe-
matician
is
a particular advantage in presenting the model and its implications to
a not necessarily numerate systems manager.
There has been considerable research into computer simulation models as
problem solving devices in many disciplines in the last decade, and it is claimed
that the sophisticated tool developed provide the social and behavioural sciences
with the equivalent to the laboratory experiments of the natural
sciences.2,3
Con-
siderable attention has been directed to simulating aspects of organizations, and
some to developing simulations of total systems. Generalized descriptions of
industrial organizations state that a firm plans by making forecasts,
operates
by
setting prices and outputs, incurring costs and administrative expenses, and mak-
ing sales. It
controls
these activities by using standards and quotes.4
There is a general lack of standard performance measures in a library system,
as with other service-oriented organizations, and consequently an inability to
make detailed forecasts of what it expects to achieve in future periods. The most
we can
say
with any certainty about
its
operation
is
that it incurs
costs
and adminis-
trative expenses, and receives funds from its parent organization.
SIMULATION AND LIBRARY SYSTEMS
Some progress has been made in using Operational Research techniques for
evaluation and research in library systems.5,6,7 These
studies,
together with earlier
systems analyses, provide useful background information for those interested in
simulating library
processes.
We were particularly stimulated by the creative use
of computer techniques as aids in systems analysis and design by the Scandina-
vians,8
and by Teichroew9 in America, and by a series of Research Workshops
on information systems at the London School of Economics. We found experi-
ence with the University of Lancaster simulation game developed from their
work on variable loan and duplication policy10 particularly helpful to the learning
process, and reports of simulation studies in American information and library
systems encouraging.11'12'13
A simulation project starts with the creation of a model of the system which it
is
required to reproduce. A non-intrusive model structure, a conceptual frame-
work that describes reality in a language readily understandable to the eventual
user of the information provided by the model, is a necessary foundation for the
acceptance of the model's validity, and the subsequent implementation of recom-
mendations based on this information. When the description of the working of
2
To continue reading
Request your trial