UNFADING DECISIONS
A Position Paper on Decision Reconstruction
Francisco Antunes
Institute of Computer and Systems Engineering of Coimbra, R. Antero de Quental, 199, Coimbra, Portugal
Management and Economics Department, Beira Interior University, Estrada do Sineiro, Covilhã, Portugal
João Paulo Costa
Institute of Computer and Systems Engineering of Coimbra, R. Antero de Quental, 199, Coimbra, Portugal
Faculty of Economics, Coimbra University, Av. Dias da Silva, 165, Coimbra, Portugal
Keywords: Decision Reconstruction, Decision and group support systems, Knowledge management.
Abstract: The importance of understanding and recording past decisions increases when we realize that employees’
memories are not always available, neither will they last permanently in organizations. In this paper we
posit that the ability to perform decision reconstruction using a Group Support System (GSS) can provide a
flexible solution to the problem, but only if the information model underlying it is able to provide
bidirectional support to the phases of a decision-making process. For this, we present general characteristics
for an information model to support decision-making as well as decision reconstruction processes.
1 INTRODUCTION
Probably there are many occasions when the sole
review of discussion topics and resulting decisions is
enough to recall the details of the decision process,
especially if the people who review them are the
same decision agents who were involved in it. Still,
as those decision agents may no longer be in the
organization, we believe that anyone should be able
to retrieve that information easily. In these
circumstances, the GSS, whose features (which are
described in detail, for instance, in Bafoutsou &
Mentzas, 2002) should allow in-depth examination
whenever required. In addition, GSS constitute a
technical element for organizational memory (as
defined, for instance, in Ackerman, 1998; Hoffer &
Valacich, 1993; Lehner & Maier, 2000; Stein &
Zwass, 1995; Walsh & Ungson, 1991) and decision
reconstruction (DR).
Methodologically, our research lies within the
scope of design research. This option takes into
consideration the creation, use, study and
performance evaluation of artefacts in order to
understand, explain and improve information
systems (Hevner, March, Park, & Ram, 2004; March
& Smith, 1995). We have adopted the process
defined by Peffers (explained in Peffers, et al., 2006)
because it is an eclectic approach, which combines
the research steps of other authors and it emphasizes
knowledge use and development, throughout the
research. We tested initial ideas on decision
reconstruction through laboratorial tests and case
studies (published elsewhere). From the gained
insights we found that most of our considerations
were ratified, but there were still unaddressed issues.
This paper presents the combination of all our
findings, whose discussion will, hopefully, gather
extended insights, before performing a second
testing round (iteration).
2 DECISION RECONSTRUCTION
We define decision reconstruction as the process that
allows an individual or group of individuals (the
decision reviewers), whether internal or external to
the organization, to understand how a GSS
supported group has reached a previous decision.
As stated in the introduction, GSSs are a natural
solution for distributed collaborative of work,
providing structured opportunities to engage in
deliberative exploration of ideas, evidence and
374
Antunes F. and Costa J..
UNFADING DECISIONS - A Position Paper on Decision Reconstruction.
DOI: 10.5220/0003080503740377
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2010), pages 374-377
ISBN: 978-989-8425-29-4
Copyright
c
2010 SCITEPRESS (Science and Technology Publications, Lda.)
Table 1: Support needs.
Decision process Reconstruction process
Argumentation
Cover a multiplicity of argumentation models.
Express the relationships among the argumentation elements.
Maintain and evidence the linking scheme of the used argumentation
model.
Structure
Create meaningful categories.
Register information evolution in time.
Link information element between discussions.
Review of the in-between steps of a decision process.
Turn information elements into an “inactive” state, instead of its deletion.
Decision-making
Use computer-guided decision-making techniques
Use manual convergence methods
Access the details of the performed convergence processes.
argument (Osborne, 2010).
Nevertheless, as GSSs are built upon the idea of
cumulative (sequential) support for the decision-
making phases (as defined by Simon, 1977, i)
intelligence phase; ii) design phase; and iii) choice
phase), it is not always easy to understand the earlier
stages of a discussion. This is particularly evident at
the end of discussions when classes, which were
created to encompass the discussion elements and
some of the details, are “flattened”.
Understanding how the past decisions affect
present ones fosters the relationships between
information and facilitating the use of knowledge in
mutually dependent contexts (Guerrero & Pino,
2001). We believe that by fostering the decision
reconstruction ability of GSSs, we promote their
capability for information retrieval, thus contributing
to ease and deepen the comprehension of past
decisions, while fostering knowledge acquisition. In
addition, expanding GSSs capabilities from the
perspective of knowledge management can
significantly improve the performance and
satisfaction of group meeting participants (Hung,
Tang, & Shu, 2008). We also stand that decision
reconstruction can enhance transparency (as stated
in Danielson, Ekenberg, Grönlund, & Larsson, 2005;
Stirton & Lodge, 2001), and will empower GSSs as
tools for public consultation and the external
scrutiny of decisions.
It is known that GSS solutions should cover a
multiplicity of approaches to support different ways
of building a collaborative discourse (according to
Turoff, Hiltz, Bieber, Fjermstad, & Rana, 1999).
These ways range from a simple question-reply
pattern to more elaborate argumentation models
supported by argumentation theory (as seen, for
instance Bentahar, Moulin, & Bélanger, 2010; Kunz
& Rittel, 1979; Maleewong, Anutariya, &
Wuwongse, 2008; Toulmin, 2003). A general GSS
information model for decision reconstruction needs
to be able to register (document) the in-between
steps of the convergence/consensus-building
provided by the interconnection of the
argumentation elements presented by the group,
during the discussion. This type of behaviour
resembles the capabilities of entity-based versioning
systems, which can create versions of packages,
classes, and even individual methods of a complete
system over its entire lifespan (Robbes & Lanza,
2005). The fine-grained ability to version
argumentation elements allows its in-depth
registration and to evidence their evolution over
time.
Another issue in decision reconstruction regards
the validity of the organizational memory. When
information expires (whether based on
administrator’s decisions or determined by existing
laws), a cleaning process can occur. We stand,
however, that the deletion of such information might
constitute an important barrier to decision
reconstruction, even when earlier information is
“flattened” to some condensed form. To this matter,
no records could mean no memory and,
consequently, the inability to retrieve past decisions.
Having in mind the intention to register all the
steps in decision-making to foster decision
reconstruction, instead of deleting information,
contributions should be marked as “active” or
“inactive” in order to be considered in the group
analysis (meaning that an inactive contribution
represents a “deletion” but without actual
information loss). We stand that it is possible to
embed the previous characteristics into an
information model to support both decision-making
and decision reconstruction, by incorporating three
different, though implicitly intertwined, types of
Table 1.
3 BUILDING THE DECISION
RECONTRUCTION SUPPORT
In order to develop an information models that fits
the abovementioned needs, whether in decision-
making or in decision reconstructing, previous
research (published elsewhere) makes us stand that
UNFADING DECISIONS - A Position Paper on Decision Reconstruction
375
is necessary to: create a flexible support for inserting
group’s contributions; link the contributions;
establish associations among information elements;
and ensure the recording of the in-between steps of
the group meeting.
A flexible model to support a wide range of
argumentation models should address contributions
as independent elements, without imposing any sort
of pre-established associations. The connection of
contributions should provide the support to establish
void links, and a different support to characterize
such links, as there are different types of expressed
connections. These connections depend on the:
argumentation model relationships (e.g. support,
response to, evidence for, etc.); structuring support
(as one of the most common features in GSS is their
ability to separate contributions into meaningful
categories or information containers, namely,
discussions, topics, categories, information
“buckets”, documents, etc.); and time-span
association (sequence, dependence, versioning,
merging, etc.). The creation of a void connection
network creates the possibility to develop a multiple
characterization framework that does not have to
impose any type of relationships, structure or
argumentation scheme beforehand.
Depending on the discussion, decision-making
support might benefit from the use of formatted
contributions or from predefined data-types used
when inserting data, especially when quantitative
data is under analysis (e.g., percentage numbers,
weights, etc.). Therefore, the connection support
could also address the data validation rules over
contributions, in order to ease or automatically
support later convergence processes. As different
discussions (or discussion segments/phases within
discussions) may require distinct argumentation
schemes, it is important to offer the support for
using different argumentation models, as usual GSS
embed an athwart representation for the whole
discussion.
The connection among contributions requires
additional characterization to define their
argumentation role within the GSS, but such
characterization should not be embedded within the
contribution support, but using associated meta-data.
In addition, different types of connections should
also be expressed using meta-data. Such association
types must include: argumentation model
relationships; structuring support; and time-span
association.
Expressing more complex argumentation models
as simpler ones does not seem troublesome. The
opposite, however, may not be accomplishable (at
least automatically) due to the lack of associated
information. Producing such information requires
the establishment of new types of associations)
beyond the ones established in the decision process.
We believe that two processes (or their
combination) could be tried, in order to achieve the
desired situation. The first one would be the
reviewer’s manual supply of the relationship
properties as individually perceived. To support this
process, the GSS should ask the reviewer to input all
the necessary association attributes, according to the
intended argumentation model. The second
procedure could use automatic mechanisms, i.e.
intelligent agents, to perform a semantic and
syntactic analysis of the different contributions and
propose the type of detected relationships to be
confirmed by the decision reviewer.
The capture of the relationships between the
discussion elements covered by the information
model should also provide the necessary basis for its
visual representation. In order to enhance its utility
in decision reconstruction and especially to respond
to different information needs and cognitive styles of
decision reviewers, it requires, nevertheless, a
combination with tools for filtering, sorting,
selecting and displaying multiple relationships.
When supporting groups in achieving decisions,
divergent contributions may exist. To deal with this
situation a GSS should provide converging and
decision-making techniques. However, as achieving
the final decision might require more than one
convergence process and more than just one
convergence method (whether manual or computer-
guided), GSS should provide a versioning capability
over the used argumentation elements and
convergence processes. Maintaining a record of the
convergence process, as well as the used methods,
contributes to ease the decision reconstruction
processes by saving and linking the in-between steps
of the decision process. The in-between recording
should also allow the production of better
reports/documents derived from the GSS decision
processes, because usual reports only embed the
latest result, especially when reporting is an
automatic feature.
Any decision report should encompass the
reasons that explain the decision outcome. However,
the process that selects such reasons and its
relevance is not a standard or an always-clear one.
As decision reviewers might not share the
relevance pattern or judgment assessment expressed
in the produced documentation, decision
reconstruction might be hindered. It would be
interesting if a GSS could parameterize automatic
KEOD 2010 - International Conference on Knowledge Engineering and Ontology Development
376
recording procedures (coarse or fine grained), in
order to produce a final document or report, for
instance, based on the performed convergence
processes, which recorded the decision evolution
within a certain time-span.
4 FUTURE RESEARCH
This paper presented the combination of initial ideas
on decision reconstruction and gained insights from
earlier testing, from which we have outlined what
we posit to be the set of fundamental characteristics
to develop an information model to support the
decision-making process, as well as the decision
reconstruction process. The defined methodology
dictates the need for a second testing round where
we intend to embed the proposed characteristics into
a GSS prototype and to submit it to further
laboratorial evaluation and case study analysis.
ACKNOWLEDGEMENTS
This work is encompassed within the aims of the
FCT PTDC/EGE-GES/113916/2009 project.
REFERENCES
Ackerman, M. S. (1998). Augmenting Organizational
Memory: A Field Study of Answer Garden. ACM
Transactions on Information Systems, 16(3), 203-224.
Bafoutsou, G., & Mentzas, G. (2002). Review and
functional classification of collaborative systems.
International Journal of Information Management, 22,
281-305.
Bentahar, J., Moulin, B., & Bélanger, M. (2010). A
Taxonomy of Argumentation Models used for
Knowledge Representation. Artificial Intelligence
Review, 33(3), 211-259.
Danielson, M., Ekenberg, L., Grönlund, Ä., & Larsson, A.
(2005). Public Decision Support - Using a DSS to
Increase Democratic Transparency. International
Journal of Public Information Systems, 1, 3-25.
Guerrero, L. A., & Pino, J. A. (2001, 7-9 Nov.).
Understanding organizational memory. Paper
presented at the XXI International Conference of the
Chilean Computer Science Society, Punta Arenas,
Chile.
Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004).
Design Science in Information Systems Research. MIS
Quarterly, 28(1), 75-105.
Hoffer, J. A., & Valacich, J. S. (1993). Group Memory in
Group Support Systems: A Foundation for Design. In
L. M. Jessup & J. S. Valacich (Eds.), Group support
systems: new perspectives (pp. 214-229): Macmillan
Publishing Company.
Hung, S.-Y., Tang, K.-Z., & Shu, T.-C. (2008). Expanding
Group Support System Capabilities from the
Knowledge Management Perspective. Journal of
International Technology and Information
Management, 17(1), 21-42.
Kunz, W., & Rittel, H. (1979). Issues as Elements of
information Systems. Unpublished Working paper No.
131. Studiengruppe für Systemforschung, Heidelberg,
Germany.
Lehner, F., & Maier, R. K. (2000). How Can
Organizational Memory Theories Contribute to
Organizational Memory Systems? [Kluwer Online
Journals]. Information Systems Frontiers, 2(3), 277-
298.
Maleewong, K., Anutariya, C., & Wuwongse, V. (2008).
A Collective Intelligence Approach to Collaborative
Knowledge Creation
Paper presented at the Fourth International Conference on
Semantics, Knowledge and Grid (SKG), Beijing:
China.
March, S. T., & Smith, G. F. (1995). Design and natural
science research on information technology. Decision
Support Systems, 15, 251-266.
Osborne, J. (2010). Arguing to Learn in Science: The Role
of Collaborative, Critical Discourse. Science, 328,
463-466.
Peffers, K., Tuunanen, T., Gengler, C. E., Rossi, M., Hui,
W., Virtanen, V., et al. (2006, February 24-25). The
Design Science Research Process: A Model for
Producing and Presenting Information Systems
Research. Paper presented at the First International
Conference on Design Science Research in
Information Systems and Technology, Claremont, CA.
Robbes, R., & Lanza, M. (2005). Versioning Systems for
Evolution Research. Paper presented at the Eighth
International Workshop on Principles of Software
Evolution (IWPSE'05).
Simon, H. (1977). The New Science of Management
Decision. Englewoods Cliffs, NJ: Prentice Hall.
Stein, E. W., & Zwass, V. (1995). Actualizing
organizational memory with information systems.
Information Systems Research, 6(2), 85-117.
Stirton, L., & Lodge, M. (2001). Transparency
Mechanisms: Building Publicness into Public
Services. Journal of Law and Society, 28(4), 471-489.
Toulmin, S. (2003). The Uses of Argument - Updated
Edition. Cambridge: Cambridge University Press.
Turoff, M., Hiltz, S. R., Bieber, M., Fjermstad, J., & Rana,
A. (1999, January 5-8). Collaborative Discourse
Structures in Computer Mediated Group
Communications. Paper presented at the 32nd Hawaii
International Conference on System Sciences
(CD/ROM), Hawaii.
Walsh, J. P., & Ungson, G. R. (1991). Organizational
Memory. The Academy of Management Review, 16(1),
57-91.
UNFADING DECISIONS - A Position Paper on Decision Reconstruction
377