=Paper=
{{Paper
|id=Vol-552/paper-7
|storemode=property
|title=Knowledge Federation as a Principle of Social Organization of Knowledge Creation and Sharing
|pdfUrl=https://ceur-ws.org/Vol-552/Karabeg-Lachica-KF08.pdf
|volume=Vol-552
}}
==Knowledge Federation as a Principle of Social Organization of Knowledge Creation and Sharing==
https://ceur-ws.org/Vol-552/Karabeg-Lachica-KF08.pdf
Knowledge Federation as a Principle of Social
Organization of Knowledge Creation and Sharing
Dino Karabeg1 and Roy Lachica2
1
Deptartment of Informatics, University of Oslo, Pb 1080, 0316 Oslo, Norway
dino.karabeg@ifi.uio.no
2
Bouvet AS, Pb 4430 Nydalen, 0403 Oslo, Norway roy.lachica@bouvet.no
Abstract. We introduce knowledge federation as a principle of social
organization of knowledge creation and sharing that is conducive to improving
our collective intelligence (ability to understand and handle complex issues).
We point at the advantages of knowledge federation and propose a strategy to
realize it in practice.
Keywords: knowledge federation, knowledge organization, collective
intelligence, socio-semantic web.
Dedication. We dedicate this article to Doug Engelbart in recognition for his guiding
insights and inspiration.
1 Introduction
Doxa is the term used by sociologist Pierre Bourdieu for a peculiarity of our
perception, due to which our existing social order appears to us as equally immutable
as the natural one. Unlike orthodoxy which, while allowing for only one ‘right’ order
of things to exist, acknowledges the existence of other ‘deviant’ possibilities, doxa
denies even the possibility of other options. We begin this article with this word,
because doxa highlights a blind spot in our perception of social reality behind which,
we submit, extraordinarily large possibilities for creative contribution are hidden.
Doxa is a reason why we try to solve our problems by thinking as we did when we
created them. And why we create new technology to support the patterns of usage that
developed based on the old technology. While the examples are as ubiquitous as the
computer desktop and email, we here focus on a much larger example – knowledge
creation and sharing as a whole.
The following two vignettes will help us introduce our theme.
1.1 Fragmentation of Sociology
After the Second World War sociology grew dramatically, and by the 1980s the
number of sociologists and sociology publications increased more than five-fold. At
the same time, sociology divided itself into a number of regional and methodological
sub-specialties, which were rapidly losing contact with one another.
The disadvantages of this style of organization were easily recognized, and in the
1989 a conference was organized by two leading researchers, European Pierre
Bourdieu and American James Coleman, to explore the possibility of bridging the
dividing lines and putting sociology back together. In the epilog to the book that
resulted from this conference, Bourdieu argued that “the progress of scientific reason
in sociology hinges crucially on a transformation of the social organization of
scientific production and communication.” [1]
Bourdieu’s argumentation is insightful and worth quoting:
“Max Weber (1978) reminds us that, in the art of warfare, the greatest progress
originated not in technical inventions but in transformations of the social
organization of the warriors, as for instance in the case of the invention of the
Macedonian phalanx. One may, along the same line, ask whether a
transformation of the social organization of scientific production and circulation
and, in particular, of the forms of communication and exchange through which
logical and empirical control is carried out would not be capable of contributing
to the progress of scientific reason in sociology—and to do so more powerfully
than the refinement of new technologies of measurement or the endless
warnings and ’presuppositional’ discussions of epistemologists and
methodologists. I have in mind here a scientific politique—that is, policy and
politics—whose goal would be to foster scientific communication and debate
across the many divisions associated with rational traditions and with the
fragmentation of social science into empirical subspecialties, theoretical
paradigms, and methodological schools.” [1]
The same reasoning may be taken a step further. While Bourdieu’s focus was on
the progress of scientific reason in sociology, the disadvantages of the fragmented
organization of sociology become more accented when we consider them in the
context of the knowledge production in the society. A consequence of the
fragmentation of sociology is that our society no longer has the sociology to inform it
about its problems.
The Club of Rome was organized to supplement this vitally important role.
1.2 Institution of the Club of Rome
The Club of Rome is an international think tank, instituted in 1968 by business
leader and humanist Aurelio Peccei and scientist Alex King, with the mission to
develop reliable insights about the risks and prospects of our global society in rapid
change. While the awareness of the global risks is now common, this was not the case
in 1972 when the Club voiced its alarm [2]. From their rich history, we shall highlight
only an insight that was made already on their initial meeting in Rome, by Erich
Jantsch in his keynote speech, where he argued that the global social system is lacking
feedback (mission-critical knowledge), and therefore also control (sustainability).
The organization and the activity of the Club of Rome contrasted sharply with the
organization and the manner of development of academic sociology. The Club
insisted that it was impossible to understand or solve any of the characteristic global
problems by focusing on the problem, and that we needed to focus on what they
called the world problematique, the condition that encompasses them all. To that end
a trans-disciplinary organization was needed, and the Club of Rome organized itself
accordingly, to include as much as possible all disciplinary and regional points of
view.
It may be interesting to observe that since its inception, the Club of Rome
struggled with the prevailing mind set in the academia. Summarizing the experiences
of the constitutional meeting in Rome, Peccei wrote:
”We discussed how the problematique should be approached in the context of
the world system and wanted to share our views with a number of scientists,
economists and sociologists. [...] The rationale behind this initiative was that, if
some ten or so Europeans of different origins and disciplines can be made to
agree on no matter what among themselves, then even the devil can be
reconciled with holy water. [...] For instance, a controversy arose over the
difference between the word ‘system’ in English and the French cognate
‘système’, spawning a whole hour of sophisticated debate [...] After 2 days it
proved virtually impossible to have the conferees agree among themselves, not
even on mere prolegomena.” [3]
1.3 Preeminence of Social Organization
We may now paraphrase Bourdieu by saying that “Progress hinges crucially on a
transformation of the social organization of knowledge production and
communication.” We have omitted the qualification “of scientific reason in
sociology” because (1) the described fragmentation of sociology is an instance of a
general trend and (2) a different organization of knowledge work and even motivation
for knowledge work are needed to secure progress, as the second vignette illustrated.
When Bourdieu voiced his appeal, there was no Web, no Wikis and no Topic
Maps. With the book and the classroom as media, perhaps a radically different social
organization was not yet possible. Now the technology allows us to implement just
about any form of social organization we may be able to imagine.
1.4 D’Après Bourdieu Project
The question remains ‘What can we do to facilitate the development of a remedial
social organization of knowledge creation and sharing in practice?’
As our knowledge federation research community is convening for the first time,
and is about to define its problem space and its manner of working, we propose to
adopt the above question as our main focus of interest. We propose to organize the
work in our community as a cooperative strategy game, where our shared goal will be
to discover ‘moves’ that can further the above cause. We call this line of work
D’Après Bourdieu (According to Bourdieu) Project, to remind us that, according to
Bourdieu, it is likely to lead to greatest progress.
Notice that the task of developing a better social organization has a self-referential,
Catch 22-like structure: The effects of whatever we might do or say might end up
being confined to our community. The reason is that there is little or no
communication between the communities – the very problem we have undertaken to
solve.
We propose to handle this difficulty by creating a small instance of a remedial
social organization involving researchers from a variety of communities, who will
make the task of implementing a remedial social organization their shared goal. Or in
other words, we propose to use ‘bootstrapping,’ a general strategy developed by
Douglas Engelbart [4].
We submit that this strategy – evolving an instance of a better social organization
of knowledge production and sharing – will profile us with respect to other similarly
motivated communities and efforts, such as the Global Sensemaking and the Semantic
Web. It will also allow us to synergize with their work, and to augment their prospects
of success.
1.5 Organization of this Article
The remainder of this article is organized as follows. In the second section, we define
and motivate knowledge federation and our proposed strategy. One of the key
advantages of taking the systemic approach by placing the social organization and the
development of an instance of the whole system into the center is that this approach
allows us to see what building blocks may be needed, and to use this to orient
research. In the remaining three sections we describe three such building blocks: the
Domain Map Object, the Value Matrix Object and the Knowledge Federation
Methodology. In the concluding section we generalize our discussion and point at the
proposed research agenda as a way to develop a methodology for ‘programming the
Web.’
2 Why Knowledge Federation
We explain knowledge federation by comparing it with two familiar principles of
organization – conventional authoring of documents, and Wikipedia-style knowledge
unification. We then discuss the advantages of our proposed strategy.
2.1 What is ‘Knowledge Federation’
The conventional practice – authoring documents, without any systemic effort to
organize the documents and the knowledge they contain into a coherent system, has
the disadvantage of leaving the following Herculean tasks to the reader:
• Identifying and assembling the knowledge resources that are required for
making a qualified judgment about a subject.
• Reconciling the divided or even incompatible opinions in order to arrive at
an informed conclusion.
• Keeping the knowledge up to date and securing that it reflects the state of the
art.
As the number of authored documents increases, these tasks tend to become
impossible. To cope, the knowledge workers specialize and divide their domains,
which leads to the problem described in Introduction.
An alternative is Wikipedia-style knowledge unification. In the Wikipedia, there is
only one article associated with any subject. That article represents (an approximation
of) the consensus position of the global community about its subject and reflects the
current state of knowledge.
This principle of organization, however, also has disadvantages:
• The simple social process by which the consensus is reached (overwriting)
obviously leaves room for improvement.
• It is difficult to secure that the authors receive credit for their work.
• It is difficult to show conflicting or contradictory points of view when those
exist and need to exist.
Recognizing these limitations, Wikipedia disallows the publication of the results
that have not already been published via conventional channels. But this means that
the Wikipedia-style organization is suitable only for an encyclopedia, and not as an
alternative to the conventional practice in the sciences and the media.
We define knowledge federation as the principle of organization that combines the
advantages of the above two approaches and avoids their disadvantages (we italicize a
concept when we attribute to it a local meaning, i.e. the one defined in this article). If
we liken the conventional authored documents to independent states, and Wikipedia-
style knowledge unification to everyone being coalesced into a single state, then
knowledge federation may be likened to political federation, which aims to reconcile
the demands for autonomy of the local units with the interests of the larger whole they
compose together.
We let knowledge federation point at the broad spectrum of social organizations
that are possible between the conventional document authoring and the Wikipedia-
style knowledge unification as simple extreme points.
Knowledge federation is both a social organization of knowledge production, and
an organization of knowledge resources. Used as a verb, it denotes the corresponding
activities.
2.2 Advantages of the Proposed Strategy
Since in the academia we are accustomed to considering an analytical result
(observing how the world is) as a contribution, and everything else as a means to that
end, or else as a distraction, the wisdom of focusing our attention at the practical
development of knowledge federation may be questioned. We offer the following
arguments to support this orientation:
• Global brain argument. Think of the totality of the people (researchers,
journalists…) and social processes that create knowledge as the global brain,
whose vitally important role is to provide our ‘social organism’ awareness
and guidance. As we have seen, the fragmented organization does not yield a
sort of global brain we can rely on. If now the global organism appears to be
acting unintelligently or even self destructively, should we not begin the
remedial action by examining what is likely to be the root cause of this
problem? Should we not give the task of inter-connecting our global brain,
and organizing it intelligently, a highest priority on our sustainability
agenda? We propose our strategy as a way to bootstrap the self-organization
of the global brain.
• Technology adoption argument. The existing social organization of
knowledge production and sharing (book and article document formats, peer
reviews, journals, promotion criteria…) is a complex system that has
evolved during several millennia, based on the book and the lecture hall as
media. To highlight its complexity, imagine it as a transportation system, and
imagine a different social organization, the one that would be capable of
truly taking advantage of the possibilities that the technology now has to
offer (multimedia, Semantic Web, wikis…) as an entirely different, non-
polluting transportation system that would offer the combined advantages of
the automobile and the public transportation. Then our current way of
deploying the technology would resemble bringing to the market various
building blocks of this new transportation system (engines, transmissions,
steering…), produced by different companies, who never secured that their
products fit together into a meaningful whole, and who don’t even know
what the whole thing is supposed to look like. Our proposed strategy would
be like creating a prototype instance of the whole transportation system. By
doing that we enable large-scale adoption of the existing technology, and
facilitate the creation of the components that are still missing.
• Efficiency argument. To see the possibility of orders-of-magnitude
improvement in efficiency and effectiveness, imagine a federated university
course, where the learning resources are co-created by international
contributors and offered to learners worldwide. Instead of having to create an
entire textbook or lecture slides, an instructor is able to focus on a single
lecture or part of a lecture, and also join forces with creative video artists,
animators and communicators, who are themselves members of the
federation. The learning resources are created and kept up to date by the
people who have the best knowledge of the material. The learners too
participate in the creation, evaluation and ‘digestion’ of the knowledge
resources, and in that way complete a well-functioning knowledge
ecosystem.
• Contribution to knowledge argument. Normally, any of us is capable of
contributing only one (fictitious) person-lifetime amount of knowledge. But
if we manage to improve the over-all system of knowledge production, and
thereby augment everyone’s ability to contribute and acquire knowledge by,
say, only 1%, this would have an effect comparable to millions of personal
contributions. And as we have seen, a much larger, qualitative improvement
may be possible.
The main insight from the punctuated equilibrium theory of evolution may be used
to complete this argument, by explaining why our proposed strategy (producing a
small functioning instance of knowledge federation within our own community) is a
natural way to approach the above goal (of affecting the global system): While
affecting the large system directly may be difficult or impossible, evolving a new
species in a fragment of the population, ‘isolated at the periphery of the ancestral
range’ is the way in which evolution can make a ‘jump.’ And that is exactly what we
are proposing.
3 Domain Map Object
We make initial progress towards a knowledge federation system design by
identifying its three large building blocks. We begin with the domain map object.
3.1 Geographical Maps, Topic Maps and Domain Maps
The main purpose of the Domain Map Object is, intuitively, to provide a ‘map’ for
placing and locating knowledge resources.
A reader familiar with Topic Maps will need no introduction to the usefulness of
such maps. Like a topic map, the Domain Map will allow for a subject-centric
organization of knowledge resources, where each subject points at all the resources
that are about that subject. To properly make an article or any other knowledge
resource known and accessible, its author will not only publish it, but also ‘place it on
the map.’
A significant difference between a topic map and a domain map is that the former
is a topical index into a domain, while the latter represents a domain. In this regard a
domain map is more similar to a geographical map. A domain map should not only
map the existing knowledge, but also orient research, by showing the areas where
knowledge is still lacking.
To that end, the Domain Map Object will provide functions for co-creating abstract
views of a domain.
Multiple views of a domain will be provided (analogous to geophysical, political,
climatological and other maps in geography).
The representation of domain maps invites creative use of visual techniques.
3.2 Exports and Imports
In addition to serving as a map of a domain, the Domain Map Object will also serve
for communication between domains.
To that end, the Domain Map Object will have provisions for importing knowledge
from other domains, and for exporting knowledge to other domains.
4 Value Matrix Object
Even when all the resources are placed on the domain map, the problem of the
cognitive overload will not be solved, because there will still be millions of resources
associated with many of the subjects. A purpose of the Value Matrix Object is to
provide additional information about the knowledge resources that will enable us to
prioritize the ones that are most suitable for a given query.
Another purpose of the Value Matrix Object is to make it possible to evaluate the
contributions of the authors.
4.1 Accumulating Value Information for a Resource
The Value Matrix Object is an object associated with a knowledge resource, whose
purpose is to accumulate the information about the value of the resource, throughout
the lifetime of the resource [5].
A value matrix may be envisioned as a matrix whose rows are criteria (relevance,
quality, reliability, level of expertise needed for understanding etc.) and whose
columns are distinct ways of making valuations (expert judgment, popular vote,
number of accesses etc.).
A challenge to which the Value Matrix Object aims to provide an answer is to
maintain a record of all the data that may later be useful for estimating the priority of
a resource with respect to a particular query.
Here are some use cases:
• The user may want to access only the ground-breaking results in an area.
• The user may look for a most authoritative survey for general audience.
• The user may trust a specific researcher as an expert with profound
understanding of an area and ask for only those resources to which that chosen
researcher has given the highest recommendation.
4.2 Evaluating an Author’s Contributions to Knowledge
The habitual practice, to evaluate the academic contributions by the publication
record alone, tends to discourage the authors from spending time on knowledge
organization, and encourage them to produce large volumes. Hence the conventional
evaluations support the practices that lead to cognitive overload. Our challenge is to
do the opposite.
A value matrix will allow for evaluating the author’s contributions by multiple
criteria such as:
• Publishing one or several articles that are considered as outstanding by some
of the leading experts in a field.
• Contributig to the organization and systematization of knowledge.
• Contributing to the dissemination of knowledge from one’s own to other
domains.
A promotion committee in a particular department may take any subset of these
and other criteria into account.
This function may make the Value Matrix Object a key component of a healthy
knowledge ecology.
5 Methodology
As we have pointed out in Introduction, knowledge federation may need to involve
not only new technical solutions and new patterns of organization, but also new
practices, methods and even goals and values. A purpose of the methodology is to
foster such developments.
A methodology is a written convention, subject to mutual agreement, specifying
the fundamental assumptions and methods based on which knowledge can be created
and communicated. Being a written convention, the methodology provides a rational
basis for departing from habitual ways. By making a convention, we create a rational,
tradition-independent foundation for knowledge creation.
The traditions and communities of interest can then use this foundation to combine
and exchange their knowledge.
In what follows we identify several ways in which a methodology may benefit
knowledge federation. We base our discussion on the Polyscopic Modeling
methodology prototype (polyscopy) [6,7,8].
5.1 Specifying the Goals and the Criteria
The pursuit of knowledge has, of course, an intrinsic value that is beyond pragmatic
concerns. In our contemporary condition, however, certain new kinds of knowledge
may need to be created with high priority, because the systemic sustainability may
require that, or for other reasons. A purpose of the methodology is to express this need
explicitly, and thereby initiate the development of suitable practice.
The Polyscopic Modeling methodology defines the gestalt as an interpretation of a
situation that points at an appropriate action. An example is ‘the house is on fire.’
Polyscopic Modeling identifies having a correct gestalt as a distinguishing
characteristic of ‘being informed.’
One may be absorbed in an observation of distant stars through a telescope and
ignore that his house is on fire. While science gives us certain preoccupations and
ways of looking at things, acquiring correct gestalts requires that we remain open to
new ways of looking and use and develop new ways of making sense. In polyscopy
this is made explicit by a criterion called perspective, which specifies that information
must illuminate a subject or issue from all relevant sides and angles, so that
everything that can contribute to sound judgment is exposed.
It is easy to see how this empowers knowledge federation.
5.2 Specifying the Assumptions
The conventional practice is to create knowledge by augmenting the reality picture
shared by a community (a discipline or a tradition), by using the language or the
terminology shared by the community. The Biblical myth of the Tower of Babel
illustrates the difficulty of putting the divergent worldviews and vocabularies
together.
Polyscopy handles this difficulty by postulating that information reflects not reality
but experience, and by allowing for concept definition by postulation (explicitly
written convention subject to agreement).
To see why this is helpful, consider the question that tends to divide the traditions –
the nature and the existence of God. Polyscopy would ignore the metaphysical
differences as ‘modeling artifacts’ and would focus on experience, aiming to identify
the pertinent experiences that are similar across traditions. It would define the concept
‘God’ in a way that makes it possible to answer the most relevant questions.
By explicitly stating the fundamental assumptions, a methodology can provide a
shared, tradition-independent ground for federating knowledge, without hidden
assumptions, and without giving an unfair advantage to any of the traditions.
5.3 Specifying the Methods
A methodology can specify the methods by which heterogeneous pieces can be
combined together.
Polyscopy provides three such methods. They are called information holons. An
information holon is both a whole ‘piece of information’ and a piece in a larger
whole.
In a federated organization, the books and articles may no longer be suitable basic
units of information. Information holons are a candidate alternative.
6 Concluding remarks
The Wikipedia and more generally the Web 2.0 approach have been criticized as
“disintermediating the expert.” This presents a challenge to the academia, because
much of the academic tradition has to do with expertise (acquiring expertise, being
certified as an expert, having the prerogatives to do independent research, be qualified
to teach on a certain level etc.). In this context, our knowledge federation proposal
may be considered as an academic response to some of the existing trends in Web
development.
The current practice is that the changes of the way knowledge is created and
accessed globally are made by a handful of computer programmers, simply by
creating a piece of software and making it available on the Web.
We submit that this key social function needs to become a subject of concerted
academic attention. And since the task of developing functional ways of creating and
using knowledge requires a lot more than a single type of expertise (technological,
sociological, cognitive, legal…), this task needs to be handled by a federation of
experts.
In the recently published Handbook of Research on Socio-Technical Design and
Social Networking Systems, Thomas Erickson writes the following:
“But socio-technical design is not just about designing things, it is about
designing things that participate in complex systems that have both social and
technical aspects. Furthermore, these systems and the activities they support
are distributed across time and space [and are] in constant flux. [...] this
complexity raises a number of general questions that socio-technical systems
designers will need to address. First of all, how do we represent such systems?
How do we cast a complex system into a material form in such a way that we
can reflect on it? [...] how do we carry out reflective conversations with them?
How will we go about ensuring that we ask the right questions, from the right
perspectives, in the right contexts? Perhaps, taking a cue from participatory
design (e.g., Greenbaum and Kyng, 1991), we will need to greatly expand the
range of participants involved in the reflective processes, which in turn may
require developing new sorts of design artifacts to aid in participatory
reflection.[...] how do we ensure that eventually we converge? Or do we?
Perhaps the notion that the end result of a design process is a stable product is
old-fashioned. Perhaps we’re headed towards a future of ‘permanent beta,’ in
which things are designed so that their design may continue during use, where
the leading edge of design resides not with the producers but with the users.
[...] However things turn out, it seems clear that socio-technical design will
require new methods, new tools, new participants, and new practices.” [9]
We submit that knowledge federation is a form of organization suitable for socio-
technical design, and propose the project outlined in this article as a way to develop
the required methodology.
References
1. Bourdieu, P.: On the Possibility of a Field of World Sociology. In Bourdieu, P. and Coleman,
J.S.: Social Theory for a Changing Society. Russel Sage, 1991.
2. Meadows, D. et al.: The Limits to Growth. New York, Universe Books, 1972.
3. Peccei, A.; The Human Quality. New York, Pergamon Press, 1977.
4. Engelbart, D.: ‘The Mother of All Demos.’ Fall Joint Computer Conference, San Francisco,
Dec. 9, 1968. Video excerpt http://www.sri.com/news/storykits/videos/bootstrapping.html
5. Lachica, R., Rudan, S. and Karabeg, D.: Quality, Relevance and Importance in Information
Retrieval with Fuzzy Semantic Networks. Proc. TMRA 2008.
6. Karabeg, D.: Designing Information Design. Information Design Journal 11/1, 2003.
7. Karabeg, D.: Information must be designed. Book manuscript.
8. Karabeg, D.: Polyscopic Modeling Definition. In R. Griffin et al. (Eds.): Changing the Tides.
IVLA Selected Readings, 2004.
9. Erickson, T.: Socio-Technical Design. In Brian Whitworth and Aldo deMoor (Eds.):
Handbook of Research on Socio-Technical Design and Social Networking Systems.
Hershey, New York, 2009. pp. 334 -335.