1992
Network Mapping: Development of an Operational Relationship Technique
General description of problem
- / -
Annex 4 of Visualization
of International Relationship Networks
Documents in linear text form are at present the main vehicle for communicating
insights on conditions and problems of many kinds, whether social, scientific,
environmental, philosophic, cultural or administrative. No comment is necessary
on the difficulties created by the information explosion as reflected in
the proliferation of documents "relevant" to meeting, programme
or policy concerns. At the same time, it is increasingly difficult to communicate
key insights and concerns and the nature of their inter-relationship. They
are too easily ignored or lost in the flood of information. Documents tend
to be indigestible even if available in the appropriate language. Each
person is necessarily protected by strategies to avoid "unnecessary"
reading. The value of new "insightful" papers and reports is
thus relatively limited. The insights are effectively entombed in documents
despite efforts by documentalists to provide sophisticated systems of access.
Abstracting systems and reviews of the state of the art do not respond
to the condition. Reading/attention time is an increasingly scarce resource,
particularly amongst those in a position to act on the insights. Too often
such systems provide answers to less significant questions, whilst failing
to assist users to formulate questions more appropriate to their concerns.
Mapping possibilities
"Mapping" has been used in the sense of compilations of relevant
texts into a macro-document. It has also been used as a synonym for certain
mathematical modelling techniques. Many attempts have been made to represent
system and network relationships in different types of diagram - from the
blackboard sketch to the sophisticated flow chart (1, 3) and the glossy
artwork presentation (2); All these have obscured the possibility of a
different approach whose advantages have not been fully explored.
A simple technique is required for visually recording "entities"
of any kind (concepts, problems, organizations, animal species, etc) and
the relationships between them; The initial problem does not lie in the
conception of such displays. Indeed network displays of concepts ("arrow
diagrams"), molecules ("metabolic pathways"), groups/people
("social networks", "organization charts"), species
("food webs"), etc have already been produced.
The initial problem lies in the fact that so much effort is required to
display the information in that form, that there is very little incentive
to experiment with improvements and complementary information aids. In
their present form they are classed as relatively expensive "artwork",
involving difficult decisions of positioning and balance. The design process
is very time-consuming and any feedback concerning modifications means
the process must recommenced. Feedback and experiments with alternative
are therefore not encouraged and the technique lacks the dynamism and immediacy
it might otherwise have. The maps are quickly out-dated and become indicative
rather than significant.
Proposal
It is proposed that:
1. Appropriate steps be taken to design and test a sequence of three computer
programs:
-
a. An input program whereby users can conveniently define and record:
-
- the range of entities in which they are interested and certain user-defined
characteristics (importance, type, quantity, etc).
-
- the relationships between those entities (again with user-defined
characteristics (importance, type, quantity, etc).
-
b. A program to manipulate the data and order it such that each entity/relationship
is provided with coordinates on a 2-dimensional grid in a format suitable
to drive the plotting program (c). It is in this program that display design
considerations of balance and positioning must be resolved using mathematical
optimization routines (probably already available); It should also permit
(re)definition of the coordinate system and selection of subsets.
-
c. Program(s) to portray the diagram on the basis of the coordinates
generated by the previous program (b); Three forms of output should be
considered:
-
- graph plotters, capable of producing large diagrams, possibly in
colour. (Note that appropriate software already exists)
-
- terminal displays (CRTs), whether remote or free-standing, and particularly
of the "home-computer" variety
-
- line printer, in the event that neither of the above is available
The above programs should, to the extent possible, not be hardware bound.
They should be written in a common language such as FORTRAN.
2. Efforts should be made to test the above system on large tape-based
data sets (e.g. that arising from the Yearbook of World Problems and Human
Potential).
3. Particular attention should be given to the user's perspective:
-
a. Ease of "associating" onto the input form (or device)
and objectifying a "mental map"
-
b. Ease of modifying or changing the importance of items on a map at
any stage (e.g. as a result of discussion during a meeting)
-
c. Ease of combining and separating input from several users, so that
"personalized" maps reflecting the reality of different biases
may be obtained.
4. Efforts should be made to test the above system as an aid to clarifying
the areas of concern, agreement and disagreement, whether for meetings,
programme elaboration, or general information (well charts, etc). It is
important to know how well people can use such maps as a means of communicating
their own perceptions of the complexity of a situation and comprehending
how it is reinforced or negated by other perceptions. Whether it is possible
to use such maps as a sort of questionnaire to provoke a series of feedback/modification
iterations (a mapping analogue to the Delphi method), remains to be determined.
Further work
Whilst the emphasis should be on production of a simple operational system,
possibilities (and costs) of extending and improving the system should
be investigated.
Whilst there are many attractive possibilities for hand-drawn maps, it
is strongly recommended that the investigation focus on the means whereby
such work can be facilitated through dynamic interaction with data bases
or displays extracted from them. Otherwise it will not be possible to breakout
of the inertia of the present dependence on scarce, costly and time-consuming
graphic stills.
Conclusion
This project should provide (groups of) users with a means of portraying
an overview of their area of concern with whatever detail is required on
its underlying structure.
Clearly such displays could condense and "pack" information present
in a multiplicity of reports; The displays themselves could be presented
in an interlinked pattern exploring detail reflected in a "large scale"
display (on the "atlas" principle). They would be easy to reproduce
and convenient to use either as a focus for meeting discussions, for educational
purposes or as "memory aids" on an office wall. Transparent overlays
are also an interesting possibility.
There are a multitude of uses for such a technique. A good example is the
ability to input several hundred species, their food chain relationships,
and entry points for accumulating pollutants (e.g. mercury). The technique
would be available just as home-computers reached the point of accessibility
at which people could make use of an associating aid to "sort out"
their mental maps as a basis for a more adequate response to complex situations.
Information in this form is more effective than an abstract or a verbal
summary and as such appropriate maps could provide the first "non-linear"
meeting agendas.