June 1971
Report on the Programme to Investigate and Implement
a Computer-based
Information System at UIA
A Profile
- / -
A. Progress to date.
1. Our first objective was to design separate (but related)
systems to perform the following functions:
- produce reference books (by computer typesetting ) and print out directories
- mail addressing
- improve questionnaire operations in data collection
- to improve advertising campaigns for sale of UIA publications (UIA is
94% self-financed through sale of publications)
- on a service bureau basis for other organizations (to increase their effectiveness
and to help establish the distinction between interaction of autonomous bodies
at the policy level and at the information processing level)
- invoicing of publications
- international organization and meeting data storage (of information typed
on cards) and its analysis
2. Investigation of the computer typesetting
routines and the
organization of the information in the 1.200 page Yearbook
of International Organizations showed the following:
several different publications could betypeset from the same data base (magnetic
tape)
- a calendar of future international organization meetings
- a directory of international organization periodcals
- bibliography of international organizationmeeting reports
these other publications are in fact detailed extensions of data already
present in the Yearbook
many of the addresses ( on international organizations and their national
members) would be held and updated for publication purposes and could therefore
be used for mailing (obviating the need for a separate system). This was
particularly interesting in that it mould be possible to address letters directly
to the executives of international organizations individually.
the data for publications could be held in a form which would permit most
forms of interrogation for research purposes
In this way each operation (mailing, questionnaires, mailing
sales literature, research and directory production) could be
made complementary, with operating economies in consequence,
to insure that all UIA files were as up-to-date as possible.
The system could be used to sand questionnaires automatically,
where information was lacking, to
insure comprehensive coverage
of the organizational network.
3. The key to the above operation mas the design of the file and
records. A highly-standardized, relatively simple, file was obtained by considering
all types of organization (societies, federations, conferences, persons,
periodicals, etc.) as "entities"
-
analyzing major quantifiable characteristics ( members, budget, etc.)
of aspects of the organization so that they could be adequately represented
in certain fixed-length fields (amount, amount units, frequency, frequency
units, beginning date, end date, etc.)
-
analyzing the nature of links in the inter-organizational system
so that a wide variety of links could be represented in the same manner
to facilitate analysis. This was done so that the distinction between inter-
and intra-organizational links became a matter of analysis rather than
data organization.
- ensuring that the various publications could be decoded, regenerated, and
typeset from the data base to give a text presentation indistinguishable from
that of other reference books.
4. It was recognized that strenuous efforts would have to be made to obtain
funds for this operation. Contacts were made with the following groups which
could provide or recommend funding :
-
UNESCO and other intergovernmental organizations. It became quickly
clear that the UIA project mas too multi-disciplinary to correspond to programme
priorities of the UN specialized agencies, or their departments. In addition,
most agencies had their own data processing commitments and priorities
and very often suffered from severe budgetary constraints.
-
international relations scholars. Only a few scholars
-
responded to the multi-purpose philosophy of the system, as described
in a number of specially-prepared UIA documents. Most of those contacted
appeared to be primarily interested in information systems which would give
a special place to the processing required by their own particular research.
-
foundations. Personal contact was made with several US foundations
at the beginning of 1970. This was during the foundation tax crisis. It
mas indicated that the project was too general for existing foundation programmes.
One foundation concluded that international agencies such as UNESCO would
be a more likely source of funds.
Adequate funds to cover the estimated costs mere eventually obtained in the
form of a loan from a Belgian bank on the basis of the evaluation of the probable
sales of publications produced by the system.
5. Since the economics was very important, the file mas to be held on tape.
Every effort was made in the design to insure that it could be "flipped"
into disc-oriented access modes. This is particularly significant because
only in this may could research on networks of organizations or departments
(within or between organizations) be satisfactorily be undertaken. The file
was in fact designed to hold a "projection" of a multi-dimensional
network onto sequential records on magnetic tape. This procedure constitutes
a breakthrough from the normal research procedure on organizations
which treatseach body as an isolated entity. This approach is no longer
adequate with the increase in scholarly interest in inter- and/or complex
intra-organizational relationships.
6. The above design was undertaken from July 1969 through to June 1970.
The programming and operation was to be subcontracted to the international
service bureau, university Computing Company, through" their Birmingham
(UK) office, which had special arrangements with a company, Rocappi, Ltd.,
owning computer typesetting programmes for use on the UNIVAC 1107.
The general analysis and program specification was completed (at no charge
to the UIA) by UCC prior to quoting for the programming, on the expectation
of programming costs in the region of $17.000. It was agreed that UIA should
start actual coding operations and at the same time Rocappi engaged supplementary
staff to handle the job. The final figure quoted, at the last minute in a
revised contract, was $29.000 (in addition to which there were one-off keypunching
costs of $14.500., aside from other expected annual costs associated with
the production of the Yearbook and updating the system).
This quote was considered unacceptable for the following reasons:
-
the increase in price reflected a last-minute change of policy on the
part of UCC at a time when UIA was apparently committed to the UCC/Rocappi
formula in order to prepare the 1970-1971 Yearbook within the deadline
-
the proposed contract specified that the programs mould become UCC property
after a three-year period of exclusive use by UIA.
7. At the time when negotiations mere started with UCC/Rocappi, no other
computer company in the UK or in Europe tuas doing directory production on
a service bureau basis. The above setback brought us in touch with an ICL
software house, Dataskil, which was prepared to use ICL computer typesetting
software, through an ICL Printing Division which was going to take in outside
work from January, 1971.
A major advantage of using ICL software was that it could be used either
in the UK or on an ICL machine which was to be installed in UNESCO (Paris)
in November, 1970. It was highly desirable for political reasons that the
UIA data bank should be operated on an "international" agency machine.
In addition, such a move would constitute a precedent for collaboration between
NGOs and UNESCO at the data processing level.
8. A contract mas signed with Dataskil for the production of a detailed
system specification at a price of $3.500. This was completed by November,
1970, and took the form of a 180- page report.
In the meantime, negotiations with the UNESCO Departments of Management and
Data Processing and of Social Sciences indicated that they were strongly in
favor of having the UIA data bank accessible to their own software. There
was even a suggestion that common files should be created. A processing
charge of $60. per hour was quoted.
Use of the UNESCO machine with 9-track, tapes created hardware compatibility
problems which could only be solved by using a chain of computers:
-
paper tape input to 9-track conversion (Paris) -- updating and interrogation
on UNESCO machine (Paris) -- 9-track to disc conversion (Brussels)
-
disc to typeset output (Brussels)
These problems were solved very satisfactorily. UNESCO then indicated that
due to the tax conditions under which the ICL machine had been obtained, the
French government would not permit use of it by other bodies such as the UIA.
It was, at this stage that the decision was take to make use of the 64K
1904 available to the ICL Printing Division (London) for all
operations. A very reasonable processing cost, in the region of $60 per hour,
was quoted. The programmes written would remain in the exclusive property
of the UIA.
9. During this period the proposal was made by the representatives of the
recently-formed International Standard Book Numbering Agency that UIA should
be the agency responsible for numbering all publications produced by non-UN
international organizations. The computer system, as conceived, could have
handled this.
10. On the basis of the system specification, an expected program development
and testing cost of approximately $19.000 was given by Dataskil. UIA then
signed a contract with them for detailed program specification scheduled for
completion in March, 1971, at a cost of approximately $7.000. This took the
form of a report of 340 pages.
11. On completion of the detailed program specifications, Dataskil revised
the expected program development cost to $31.000. This unexpected increase
was attributed to an "error" in the original estimate.
In these circumstances, UIA reviewed its commitment an integrated data system
covering directory production, research, and mailing. Because of the innovation
in the concept of the system, it was recognized that a certain element of
risk mas involved in getting the system "debugged". UIA was informed
that despite the increase in price, the detailed study did not reveal a corresponding
reduction in the risk, or whether the actual running time mould prove to be
economically viable
Several other factors also required consideration:
-
the ICL Print Division indicated that directoryproduction only could
be undertaken more economically and at a programming costs of $2.400. No
research or addressing facilities would be available.
-
the design of the system had been sub-optimized for research purposes
to use a more economical tape mode. ICL possessed sophisticated software
(PLUTO) which would permit processing of the network-type structures characteristic
of the network of international and national organizations which the UIA
documents.
-
extensive investment in programming mould improve the quality of UIA
information and facilitate research, but it mould drain limited funds from
badly-needed investment in the build-up of mailing lists of organizations
affected by the international level — and therefore potential purchasers
of UIA publications (so essential to its survival as an independent organization)
-
although difficult, a less-sophisticated system could perhaps be upgraded
gradually and as funds became available, thus achieving the goal with lower
risks. The Dataskil study had indicated how this might be done in general
the scholars concerned with international studies were not moving rapidly
towards the development of theoretical models which mould benefit in any
immediate future from data banks permitting analysis of
- more than narrowly-defined categories of international organizations
-
national and sub-national links to the international level inter-and
intra-organization links
- cross-disciplinary links between programme areas
-
links other than formal and contractual
as a consequence of the previous point, it would be difficult to obtain
multi-disciplinary academic support for fund applications to foundations,
which were themselves going through a difficult period with respect to funding
of international projects, and those justified for acaemic purposes intergovernmental
agencies were not in a position to support such initiatives, due to shortage
of funds,(e.g., reduction of social science budget of UNESCO for 1971-72)
and concentration of effort on the establishment of their own agency-oriented
computer applications awareness of the significance of data processing for
their future had not yet aroused international non-governmental bodies to
any degree of effective collection action
8. Alternative Strategies.
1. Directory production systems.
UIA is highly dependent on directory sales, and therefore cannot afford to
jeopardize any further its viability in this area by experimenting at its
own financial risk on expensive modifications to existing computer typesetting
systems in order to facilitate research.
The use of computer typesetting methods must therefore be judged on economic
considerations alone. Preliminary estimates show that it would take approximately
3 to 4 years of operations to cover the initial investment. This would seem
to be an unnecessary investment in the case of the English edition of the
Yearbook of International Organizations, which has been recomposed for the
Thirteenth Edition. It could be of interest for the French Edition, which
is being prepared from manuscript (note that under the new numerical ordering
of the Yearbook, all indexes for the English Edition can be
used without change for the French edition). Detailed quotes will be obtained
on the basis of the Thirteenth Edition volume.
The other volume considered for production bycomputer typesetting methods, namely the Yearbook of International Congress Proceedings,
raises similar problems. Three-quarters of the print metal (excluding the
indexes) can be held as useable from one edition to the next. This is a strong
argument against switching to computer typesetting methods. From the computer
point of view, this volume also suffers from the disadvantage of having entries
which are held in chronological order, rather than in organization number
order corresponding to that of the Yearbook of International Organizations.
Unless it was proposed to change the ordering, the computer approach would
be of little interest. The main factor suggesting that computer methods should
be used for processing this Yearbook is the possibility that the UIA will
take over the International Standard Book Numbering for international organization
publications. In this case, it would be an advantage to be able to process
these numbers at the same time as information is prepared for the Yearbook
of International Congress Proceedings. (Special lists of numbers and publications
have to be prepared as part of the ISBN Agency function.)
It is, however, possible to envisage investing extra funds at this stage
in the programming, coding, and keypunching of directory material, so that
it can be: a. integrated with the mailing/research system at a later stage
(as explained below) b. interrogated for research purposes using specially-designed
but relatively simple word-scan programmes designed by the researcher.
2. Mailing/research system.
Several strategies are open to the UIA in this area. The choice between
them is dependent upon funds available.
a. Minimum response would result in a mailing system with one record
for data coded on the organizations in guestion.Provision for_interrogation
would be minimal Programming costs mould be in the region of $1.000 unless
standard packages mere used. This system could not be used for research purposes
to any extent.
b. On the basis of the analysis by Dataskil, a system could be devised
which could hold many data records on each organization, together with address
information and a record for each of the titles in different languages. Basically
this would be the non-typesetting half of the abandoned integrated system.
Many types of research data could be stored. The file would be structured
to reflect the intra-organizational structure, the many types of links to
many types of members in each country, and the many types of contacts with
other bodies, as well as other features.
This would be used for:
-
production of mailing address labels
-
print-out of a questionnaire to organizations which could be used as
an editorial aid in the preparation of directories and to improve the
quality and quantity of information
-
interrogation for research purposes
-
storage of historical records currently on difficult- to-search card
files
This system would be contracted to a service bureau on a lowest-cost basis.
Eventual compatability with the directory system would not be considered.
This is currently the most probable alternative.
Programming costs for this system are estimated to be $ 5.000. Coding
and keypunching costs would depend upon the amount coded, but mould also be
in the region of $5.000. This would cover inclusion of all international
organizations and all UIA clients. More addresses and coding on individual
organizations mould be introduced in a second phase.
c. Several service bureaux are quoting for standard programme packages
to handle the UIA invoicing and accounting operations, which have become excessively
complex and time- consuming, particularly with the introduction of the Added
Value Tax (T.V.A.) since January, 1971. It is quite probable that special
accounting records could be inserted into the system described in (b) to benefit
from the client address data coded for it.
d. As explained in connection with the directory production, it is still
possible to envisage slightly modifying the programmes and coding of the directory
material. This would be done to facilitate interrogation, but also to permit
progressive convergence towards a system in which directory production and
mailing/research mould be integrated.
The technique would be to supply sufficient coding on the mailing/research
tape for particular features of the international organizations (e.g., members)
to permit a special programme to be able to generate the appropriate lines
of text on the directory tape as asubstitute for the equivalent appropriately
coded paragraph inserted there by conventional computer typesetting means.
Conversion modules could be created for those paragraphs which warranted
it, as soon as funds permitted the coding and modules to be completed. Dataskil
has already supplied a detailed study of the modules required.
The costs involved are as follows:
modifications to directory programmes/coding $4.500
programming costs (as for (b)) 5.000
coding and keypunching (initial) 5.000
— Total 14.500
If it is decided not to go ahead with the English edition of the Yearbook
of International Organizations by computer typesetting methods, then this
approach could be considered for the French edition.
e. A fifth strategy which can be envisaged is to go ahead with (b) or (d),
but to insure that records designed for tape can be handled by the special
disc software used for processing networks and structures. This mould mean
that mailings and interrogation on a sequentially-ordered file (for research
purposes) would be undertaken using the (b)-type programme. Any research
on networks of organizations (or listings such as "all international
organizations with HQ in France with national member organizations in
South Africa which are also members of international organizations with
HQ in the USA") would be done by reading the tape file onto discs. Dataskil
has already made a proposal on these lines, using their PLUTO system. This
appears to be unnecessarily complex, and costly.
A relatively simple alternative programme has been devised by the UIA, discussed
with representatives of computer bureaux, to take the magnetic tape prepared
by the (b)- type system and set up the organizational network on disc for
the more complex types of research. Depending upon the service bureaux, the
programming is provisionally estimated to cost between $2000 and 2500. Implementation
of this could be considered for 1972, when more coding will be incorporated
on organizations, their members, and their meetings.
|
