=Paper=
{{Paper
|id=Vol-2859/paper12
|storemode=property
|title=An Overview of Decision Support Software: Strategic Planning Perspective
|pdfUrl=https://ceur-ws.org/Vol-2859/paper12.pdf
|volume=Vol-2859
|authors=Sergii Kadenko,Vitaliy Tsyganok,Oleh Andriichuk,Aleksandr Karabchuk,Minglei Fu
|dblpUrl=https://dblp.org/rec/conf/its2/KadenkoTAKF20
}}
==An Overview of Decision Support Software: Strategic Planning Perspective==
https://ceur-ws.org/Vol-2859/paper12.pdf
142
An Overview of Decision Support Software: Strategic
Planning Perspective
© Sergii Kadenko1[0000-0001-7191-5636], © Vitaliy Tsyganok1[0000-0002-0821-4877],
© Oleh Andriichuk1[0000-0003-2569-2026], © Aleksandr Karabchuk1, © Minglei Fu2
1 Institute for Information Recording of National Academy of Sciences of Ukraine, Kyiv,
Ukraine
2
College of Information Engineering, Zhejiang University of Technology, Hangzhou, China
Abstract. The paper features a review of available software tools and systems
that utilize expert data for decision support in weakly-structured subject domains.
Existing tools are analyzed and compared among themselves from the standpoint
of implemented mathematical approaches and functions, particularly, those re-
lated to strategic planning. We outline the key trends of decision support tools
development, witnessed during the last decades. We show that existing auto-
mated decision support tools have a bunch of limitations and drawbacks. Based
on conducted analysis, we formulate relevant requirements to modern decision
support software and specific recommendations as to selection of decision sup-
port tools for strategic planning and further improvement of existing decision
support products.
Keywords: Decision-making Support, Expert Estimation Scale, Target-oriented
Hierarchic Decomposition, Strategic Planning, Resource Allocation, Scenario
Analysis.
1 Introduction: Relevance of Decision Support Tools Usage for
Strategic Planning in Weakly Structured Subject Domains
Decision support tools using expert knowledge are mostly applied in weakly structured
subject domains. These domains are characterized by a set of features, outlined in [1].
Particularly, we should mention absence of measurement units for criteria according to
which objects are compared, high level of uncertainty of information on the subject
domain, lack of determined data, complexity of formal and analytical description, pres-
ence of multiple interconnected impact factors, absence of explicitly formulated target
function. Due to these features, it is often the case that only expert knowledge and
methods of its processing allow us to get a detailed formal description of a weakly
structured subject domain and provide the decision-maker (DM) with recommenda-
tions, which are needed to make the decision better substantiated and more credible.
In this paper we are going to focus mainly on decision-making support problems,
emerging during strategic planning in the weakly-structured subject domains. Building
of strategies in such domains, usually, calls for engagement and processing of expert
Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons
License Attribution 4.0 International (CC BY 4.0).
143
knowledge. By a strategy we propose to denominate a long-term, step-by-step, con-
structive, rational, and uncertainty-proof plan. Its implementation is accompanied by
constant analysis and monitoring, and is targeted at achievement of a certain main goal.
An essential component of strategic planning in a given domain should be an adequate
model of this domain (built, for example, in the form of a hierarchy of impact factors
[2,3]). Building of such a model allows the DM to solve a set of problems, such as:
- Strategic planning itself;
- Ranking and rating of goals, projects, factors, criteria, and alternatives;
- Evaluation of relative efficiency of projects;
- Priority-setting in the DM’s activity;
- Distribution of limited resources among projects;
- Defining the efficiency of potential decision variants;
- Generating and analysis of situation development scenarios, etc.
In order to obtain maximum information, required for decision-making, it makes
sense for the DM to involve not one expert, but a group of experts in the process. Rec-
ommendations, formulated as a result of consensus between expert group members,
will be better substantiated, than advice of a single specialist. Under modern realities,
a desirable requirement to decision support (DS) tools is an opportunity to organize
expert work in remote mode (online).
High level of complexity and uncertainty of weakly structured subject domains, the
need for interdisciplinary approaches, consideration of large numbers of heterogenous
impact factors and opinions of multiple experts from different spheres (who work
online, in the general case) call for automation of DS process, particularly, during stra-
tegic planning. Modern decision support tools and systems (DSS) are intended to sim-
plify, and, at the same time, improve the process of collection and processing of expert
information, in order to facilitate its most thorough usage. If information, obtained from
experts, is complete, consistent, and detailed; if it is not distorted during collection and
processing, then recommendations for the DM, formulated on its basis, become more
credible, and respective decisions become better substantiated.
In this paper, in order to analyze the capacities of modern DSS tools, we suggest to
use a widely-approbated original technology of strategic planning [4] as basis. The
technology involves the following conceptual phases:
1) The DM formulates the main goal of the strategic plan in the given subject domain
and recruits the expert group (for example – using co-authorship networks [5]);
2) The experts hierarchically decompose the main goal into factors, influencing its
achievement. Decomposition is performed until the level of specific projects is reached.
Projects are “atomic” factors, within the competency of the DM.
3) The experts estimate relative impact of each factor from the hierarchy. The outcome
of this phase is a DSS knowledge base (KB), describing the subject domain in the con-
text of the specific main goal of the strategic plan, represented by a weighted hierarchy
graph.
4) Finding the optimal (rational) distribution of limited available resources among pro-
jects, which maximizes the strategic goal achievement degree. Calculation of this de-
gree is performed for a given moment of time and takes into consideration the current
impact of all factors and threshold values of project funding.
The technology demonstrated its high efficiency in such applications as:
144
- Planning of defense sphere development;
- building of development strategy and evaluation of efficiency of measures o
space activity and production of space equipment in Ukraine [4];
- environmental protection sphere [6];
- information operation research and recognition [7];
- building and analysis of situation development scenarios [8, 9].
Manufacturers of DS tools worldwide also use similar approaches, and this fact ad-
ditionally testifies in favor of the technology [10].
We should note that any automated DS tools have their own original intended pur-
poses and pre-conditions of development. That is why it is problematic to single out a
unified common criterion, based on which all DS software products could be compared
with each other. In this paper we try to review the features of the most well-known tools
in the context of their ability to solve the problems, that emerge during strategic plan-
ning in weakly structured subject domains (based on expert and other information).
2 An Analytic Overview of the Most Widely-used DS Tools
Due to recent expansion of DS tools usage, we are witnessing an increase in demand
for intelligent systems intended to solve DS problems. World-known DSS, developed
in recent years, include the following ones: ExpertChoice, SuperDecisions,
DecisionLens, D-Sight, Promethee, ОЦЕНКА И ВЫБОР (Estimate and choice), So-
lon, Consensus, and others.
Common characteristic features of most DSS, irrespectively of their intended pur-
pose, are as follows: a KB, a subject domain model (decision context and estimation
criteria), and a user interface [11]. We propose to focus on universal-purpose DSS,
whose mathematical ware incorporates the most popular contemporary decision sup-
port methods.
2.1 SuperDecisions DSS and Similar Products
One of the most widely used decision support methodologies today is the analytic hi-
erarchy/network process (AHP/ANP), developed by Tom Saaty [2]. It is implemented
in SuperDecisions DSS [12]. This system largely replaced another popular product,
ExpertChoice DSS [13]. The system is designed to calculate the relative efficiency
(weight) of alternatives based on multi-criteria estimation. In the general case, the hi-
erarchy of criteria includes four sub-graphs: B – benefits, O – opportunities, C – costs,
and R – risks.
Weights of each of alternatives and importance of criteria are defined through expert
pair-wise comparisons performed in fundamental scale, or through direct estimation.
Obtained estimates are aggregated through weighted summation. Estimates according
to criteria, representing benefits and opportunities are considered with “plus” sign,
while estimates according to criteria, representing costs and risks – with “minus” sign.
145
The range of applications of the system is extremely wide. Proceedings of the Inter-
national Symposiums for the Analytic Hierarchy Process for the past 32 years (1988-
2020) prove it [14].
SuperDecisions is just one of the many DSS, based on AHP/ANP methods. Other
DSS, where these methods are used, include DecisionLens, ExpertChoice,
MakeItRational, MindDecider, RationalFocalPoint (RFP), SmarterGovernment. At the
same time, complete strategic planning cycle, outlined in the introduction, is imple-
mented only in DecisionLens system (which has a license cost of 25,000-30,000 euro,
and no trial version). Other systems are focused on ranking and rating of decision var-
iants according to some aggregate criterion.
So, SuperDecisions and most of the similar systems feature only separate steps of
the strategic planning technology. DecisionLens, the only system, in which resource
allocation phase is actually implemented, is much less affordable, than other DSS, due
to high license cost. Besides that, most systems have limitations, concerning remote
expert session organization.
2.2 PROMETHEE Visual DSS
Just like SuperDecisions, this DSS is intended for ranking of several decision variants
(alternatives) based on their estimates according to several criteria. PROMETHEE DSS
and DS method of the same name (the abbreviation, translated from French, stands for
“the method for organization of preferences for enrichment of estimates”), embedded
in it, were developed and described by Belgian researchers [15]. Promethee Visual is a
“successor” of previous products, in which the method was implemented, such as
PromCalc and DecisionLab.
The estimates are aggregated through weighted summation. Object estimation crite-
ria can be both quantitative and qualitative. Estimates can be absolute or relative. The
peculiar feature of the method is the so-called “preference function” (U-shaped, V-
shaped, Gaussian, stepwise), which indicates the specificity of preference relationship,
depending on one or several determinant parameters.
The system allows users to solve only separate sub-tasks of the aforementioned stra-
tegic planning procedure. Its core functions do not include resource distribution or re-
mote expert sessions.
2.3 1000minds DSS
1000minds [16] is a tool for group decision-making in remote mode. Alternatives are
ranked based on their estimates according to several criteria. The system does not in-
clude any desktop versions or applications – all the work on decision-making support
is performed online.
The mathematics behind the system is based on PAPRIKA method (Potentially All
Pairwise RanKings of all possible Alternatives) [17]. The experts are asked to provide
ordinal pair-wise comparisons of alternatives from a given set (answer the question:
“which of the two alternatives is better?”). Often the task is to reach a compromise
146
between several criteria. For example, the expert is offered to choose a costly, but prom-
ising project, or a less promising project, that costs less. The number of criteria can be
more than two. Particular order of pair-wise comparisons is intended to minimize the
number of times the expert is addressed; many ordinal pair-wise comparisons are “re-
stored” according to transitivity rules, based on already available expert judgments. The
final result of the expert session is the ranking and the rating of alternatives.
As we can see, the system is mainly focused on the first phases of the strategic plan-
ning procedure. Resource allocation function is unavailable.
2.4 AIRM online DSS
“AIRM online” DSS implements the aggregated indices randomization method
(AIRM) [18]. The method is intended to reduce the uncertainty, emerging during esti-
mation of weight coefficients: the information on the weights of criteria and alternatives
is often provided in the form of ordinal instead of cardinal estimates, or numeric inter-
vals. This information is often insufficient to calculate weight coefficients. Weight vec-
tor is chosen based on Bayesian randomization, from among all possible values, which
can be assumed by its coordinates. It is presumed that each coordinate is evenly distrib-
uted on a certain numeric interval, and, consequently, if several objects are compared
according to several criteria, then the aggregate quality indicator Q, is also a random
variable. Its value is defined as mathematical expectation of this variable, and precision
is characterized by standard deviation of the indicator.
The system is targeted at a certain type of estimates, and allows users to solve only
a certain type of problems, i.e. alternative weight calculation (this is just one of the
phases of the strategic planning technology).
2.5 Analytica Package
The visual software package “Analytica” developed by Lumina Decision Systems [19],
uses the so-called relationship or impact charts for illustrative representation of a situ-
ation, calling for decision-making.
Mathematical background of the system is, mainly, based on statistical tools, rather
than on some specific multicriteria decision support methods. Considerable attention is
dedicated to building of probability distribution functions, that characterize alternative
decision options.
The system does not support estimate input in the form of pair-wise comparisons –
the expert is required to input measurement units, ranges, extreme values of indicators,
that influence the achievement of a certain goal.
So, a considerable advantage of Analytica package is that an expert has an oppor-
tunity to set the estimation scale in the process of the expert session. However, again,
the package implements only certain phases of strategic planning cycle.
147
2.6 D-Sight DSS
D-Sight software [20] is intended for multicriteria evaluation (ranking) of alternatives
in different qualitative and quantitative scales. An expert can design his(her) own scale,
such as “a scale of professional competence”. Just like many other DSS, the system
allows us to input estimates of alternatives and weights of criteria. The software has
many tools for visualization of expert session results, particularly, using
PROMETHEE/GAIA method and multi-attribute utility theory (MAUT).
So, just like “Analytica” D-Sight has some advantages, related to selection of esti-
mation scales, however, it does not incorporate the whole strategic planning cycle, and
does not support remote work of a group of experts.
2.7 MakeItRational DSS
MakeItRational is another AHP-based DSS [21]. The system is intended for multi-cri-
teria evaluation and selection of alternatives from a given set. Initially, a set of alterna-
tives is input. Then a global estimation criterion is formulated. It is decomposed into a
hierarchy of sub-criteria, representing both negative and positive influences (costs and
benefits). Next, alternatives are estimated according to criteria. Their estimates can be
input in the form of both numerical values and pair-wise comparisons in Saaty’s fun-
damental scale. The final result of the expert session is the rating of alternatives accord-
ing to the global criterion and its sub-criteria. The system supports exporting of expert
session results into MS Excel and automatic compiling of a document with the report
on the expert session outcomes.
Thus, just like SuperDecisions, MakeItRational, mainly, focuses on problems, re-
lated to rating of alternatives. Obtained ratings can provide the basis for further strategic
planning and priority-setting, however resource allocation functions are not imple-
mented in the software.
2.8 MindDecider DSS
Just like other systems, already listed in the paper, MindDecider DSS [22] is intended
for multicriteria estimation and rating of alternatives. The expert session is divided into
several stages, each represented by a respective operation mode: planning and design,
estimation, terminal analysis, report/log compiling. First, the main goal is formulated,
then – a set of estimation criteria and ranges of estimate values. After that, alternatives
are estimated according to criteria. After aggregation of alternative estimates according
to specific criteria their global rating (in per cents) and global ranking are compiled.
MindDecider does not incorporate the whole strategic planning cycle.
2.9 LogicalDecisions DSS
Logical Decisions DSS [23] utilizes the tools of AHP and MAUT. The system is in-
tended for multicriteria estimation of alternatives from a given set. Criterion weights
148
(«measures») are defined through direct estimation, or pair-wise comparison of alter-
natives in an arbitrarily chosen scale. Alternatives are estimated through pair-wise com-
parisons, in the arbitrary scale as well. So, Logical Decisions also incorporates just
certain phases of the strategic planning cycle.
2.10 TreeAgePro DSS
Tree Age Pro [24] is intended for evaluation of different decision variants. Possible
options are presented in the form of decision tree branches. Decision tree graph (similar
to impact diagrams from Analytica system) can include nodes of different types:
“chance” (probability node), “Markovian node”, “decision node”, “Boolean node” etc.
Again, similarly to Analytica, TreeAgePro provides a whole environment for automa-
tion and structuring of expert estimation and choice process while solving various kinds
of problems. However, the mathematical tools, providing the backbone of the system,
are mostly based on probability theory. We should stress, that TreeAgePro and Analyt-
ica packages have broad opportunities for adaptation to specific problems, but respec-
tive functions have to be implemented separately.
2.11 RFP (RationalFocalPoint) DSS
RFP (Rational Focal Point) [25] is a complex of means for automation of decision-
making support process at municipal level. Particularly, these decisions concern fund-
ing of various projects, portfolio investment, and others. Decision variants are evalu-
ated according to quantitative and qualitative criteria. The system supports both pair-
wise comparisons and direct expert estimation. Since the software is intended for mu-
nicipal-level decision support, a large number of factors, whose impacts change with
time, and opinions of multiple experts, who work in remote mode, need to be taken into
consideration. So, developers of the system (first – IBM, then – Unicom) focused on
online operation mode and on the features related to modeling of different decision
options (scenarios) in time. Thus, the opportunity of working in remote mode is a strong
advantage of the system. However, narrow intended purpose of the system limits its
functional capacity a bit.
2.12 VeryGoodChoice DSS
Very Good Choice [26] system (somewhat outdated from the standpoint of imple-
mented software solutions and no longer supported by the developers), built by
MVLsoft company, is based on a famous decision support method – ELECTRE (Elim-
ination Et Choix Traduisant la Realite – elimination and choice, reflecting the reality)
[27]. The software is an add-in for Microsoft Excel package. Its capabilities in the con-
text of strategic planning are limited.
149
2.13 “ESTIMATION AND CHOICE” («ОЦЕНКА И ВЫБОР») DSS
«ОЦЕНКА И ВЫБОР» is an internet-based software system, intended for solving mul-
ticriteria decision-making problems. These problems might be hierarchically structured
(from upper-level to lower-level criteria: Purpose of analysis – Generalized indicators
– Indicators). «ОЦЕНКА И ВЫБОР» uses different decision support methods: AHP,
value functions, ordinary weighting method, Pareto dominant analysis,
BENEFIT/COST analysis.
The final result of an expert session is the rating of alternatives, calculated as
weighted sum of the respective indicators, characterizing the alternatives. The indica-
tors may be qualitative, quantitative, or Boolean ones.
We should stress that the system is initially oriented towards remote decision support
problem solution.
Remote operation and support for different estimate types are the key advantages of
the system, while definition of an optimal strategy remains beyond the scope of its ca-
pacities. Detailed description of the system’s features and functioning is provided in
[28, 29].
2.14 “SVIR’” («СВИРЬ») Instrumental System
According to the system’s developers, “SVIR’” is an instrumental system for multi-
criteria decision-making problem solution. It satisfies such requirements as: universal-
ity, simultaneous usage of objective and subjective estimates, high-dimensionality
problem solution, autonomous problem solution, opportunity for integration with other
systems for data exchange and processing, ergonomic design, ability to evolve.
Multi-criterial choice problems are solved by the system using multi-criteria optimi-
zation and classification methods, as well as AHP. Object estimation criterion weights
are set through direct estimation, or calculated based on pair-wise comparison matrices
based on expert preferences, or equivalence of features in tables or primary features.
The results are presented in a unified colored quality scale, which can be configured
and saved. On the output, the system also allows users to obtain charts of object distri-
bution according to criteria and to aggregate estimate, as well as, analyze the contribu-
tions of criteria into the aggregate estimate (perform sensitivity analysis).
So, the advantages of the system are its support for different estimation scales and
graphic representation of expert session results. Certain phases of the strategic planning
technology (such as resource allocation) need to be implemented separately, or per-
formed by additional external modules. The system and its modifications are described
in numerous publications, such as [30, 31].
2.15 “Solon-3” and “Consensus-2” DSS
“Solon-3” DSS is a tool for decision-making support based on complex target-oriented
program (CTP) building. A CTP is a set of activities, united by a common goal and
shared resources. The key tasks performed during CTP formation are
1) main goal formulation,
150
2) defining prospective directions of its fulfillment (sub-goals),
3) selection of the most efficient means (projects) and
4) resource distribution among selected projects.
In order to complete tasks 2) and 3), one needs to rank the objects (projects, goals).
Technological decision-support process using “Solon-3” DSS [32] includes the fol-
lowing phases: decomposition of the main goal and building of a hierarchy of goals,
expert estimation of partial impact coefficients of sub-goals, calculation of relative ef-
ficiency of different ways of program implementation and formulation of alternative
projects, calculation of coefficients of projects’ impact upon the main goal, which are
used as indicators of relative efficiency of the projects. To calculate the relative effi-
ciency of the projects, the given DSS uses the method of dynamic target-oriented eval-
uation of alternatives (MDTEA). Detailed description of “Solon-3” DSS can be found
in [3,32].
The system of distributed collection of expert information (SDCEI) “Consensus-2”
[33, 6] was developed and implemented to simplify the process of group expert session
organization in remote mode. The system is a cloud solution; it allows the expert group
members to join an online session, organized to solve a specific problem, and to per-
form group decomposition of this problem. The system’s interface significantly simpli-
fies task 2) from the above-mentioned list. For resource allocation to projects additional
special software modules need to be connected to the system.
2.16 Other Decision-making Support Software Products and Applications
As existing DS software is highly demanded by representatives of different application
fields, it is constantly improved and upgraded. At the same time, new products are being
developed, in the form of both standalone solutions and add-ons or apps. For instance,
Medical Sapiens web-platform for medical decision-making [34], Total Decision soft-
ware [35], and Decision Mentor app [36] were developed just a few months ago (as of
the moment of this paper preparation). Besides, Decision Mentor is an example of one
of the latest trends – DS solutions going mobile.
Extensive assortment of DS solutions, offered by online software marketplaces, also
signifies the evolution process. For instance, Capterra platform [37] (in addition to al-
ready mentioned DSS 1000minds, DecisionLens, Analytica, ExpertChoice) features
dozens of software products and whole packages. Each of them is implemented in its
own particular way and has its own intended purpose in a specific subject domain (cus-
tomer support (Zingtree), visualization of complex group decision-making situation in
large organizations (EIDOS), and others).
We should also note that “decision support software” concept assumed wider mean-
ing in the eyes of online marketplace customers. For example, Wolfram Mathematica
software (also available through Capterra platform) is a cloud solution (in fact, a whole
environment), intended for image recognition, data visualization, and other functions,
using machine learning, neural network algorithms, data mining, and other top-notch
approaches. Yet, it is still attributed to decision support software.
Moreover, criteria, according to which the software tools are compared, are shifted
towards their commercial attractiveness. Capterra platform offers the following list of
151
product comparison characteristics: initial cost, availability of a free trial version, way
of implementation (cloud solution, web application, iOS/Windows/Android, mobile
app etc), necessity for coaching sessions for users (online tutorials, documentation,
webinars etc), and availability of tech support (weekdays only, 24/7 online support,
etc). Almost no attention is dedicated to consideration (or analysis) of specific mathe-
matical and technological solutions, providing the basis of this or that DS tool. So, it is
very problematic to consider the respective tools in comparative context, especially if
you are an academic researcher rather than a beginner-level user with a narrow-profile
specific demand.
We should stress that attribution of DS software tools and applications to specific
narrow problems and subject domains (most probably resulting from business interests)
significantly limits the opportunities for their usage. From the standpoint of users, a
DSS should be universal, and, at the same time, easily adaptable to specific applied
problems in each domain (for example, through hierarchical decomposition these prob-
lems and swift adjustment of subject domain models). Otherwise, a new separate DSS
would have to be applied for each specific subject domain or problem.
Authors of [38, 39] provide a comparative analysis of some DSS, including those,
not listed in our review. Having appended the review with information on some other
DSS [38, 39], we are able to obtain aggregate comparison data on DS software, partic-
ularly, from the standpoint of mathematical and technological solutions implemented
in them (see Table 1).
If we take just the 4 conceptual phases of the above-mentioned strategic planning
cycle as DSS comparison criteria, the comparison makes no sense: all the systems we
consider do allow users to rate alternatives, factors, and projects (incorporate 3 phases
out of 4), while only one system supports resource allocation functionality. That is why,
in order to compile an illustrative comparison table, we have selected more constructive
(specific) criteria.
In addition to opportunities for alternative estimation scale selection (pair-wise com-
parison scale or others), group estimation, and online expert session organization
(through a cloud software version), several other DSS functions also significantly in-
fluence the credibility of decisions (including strategic ones). These functions include
consideration of time lags in implementation of certain components of a strategic goal,
as well as automated analysis of sensitivity of the final decision variants’ rating/ranking
to perturbations of input data.
Our review is not an exhaustive one, however it illustrates the general trends in the
development of modern DSS, using expert data (including pair-wise comparisons and
other types of estimates in various scales).
We should note, that none of the listed systems allows its users to organize the whole
strategic planning cycle (outlined in the introduction) using expert information in re-
mote mode. For example, DecisionLens (one of the most expensive DS software sys-
tems in the world) combines the maximum number of respective functions, required for
strategic planning (including resource allocation).
152
Table 1. Comparison of functional capabilities of modern DSS
DSS name Implemented Pair-wise Time lag Sensitivity Group es- Support Cloud
methods compari- analysis analysis timation for differ- (web)
sons ent scales version
1000Minds PAPRIKA Y N Y Y N Y
AIRM Online AIRM N N Y N N Y
Analytica N Y Y N N Y
DecisionLab Y N Y N N N
DecisionLens AHP, ANP Y Y Y Y
D-Sight MAUT, Y N Y Y N Y
PROMETHEE
ExpertChoice AHP Y N Y Y N Y
Logical AHP, MAUT Y N Y Y N N
Decisions
Make It Rational AHP Y N Y Y N Y
MindDecider AHP N Y Y Y N N
PROMETHEE PROMETHEE Y N Y Y Y N
Visual
RFP AHP, ANP Y Y Y Y N Y
SuperDecisions AHP, ANP Y N Y N N N
TreeAgePro N N Y N N N
Very Good ELECTRE Y N Y Y N N
Choice
ОЦЕНКА И AHP, utility Y N Y N Y
ВЫБОР (Estima- function, Pa-
tion and choice) reto Analysis
SVIR’ Multi-criteria Y N Y Y N
optimization,
Classification,
AHP
SOLON MTDEA Y Y Y Y Y Y
However, it has its own limitations concerning online expert sessions and expert
estimation scale selection. 1000minds, RFP, “ОЦЕНКА И ВЫБОР” (“Estimation and
choice”) systems are intended for remote work of experts, but they focus on ranking
and rating of alternatives. D-Sight and PROMETHEE Visual offer wide opportunities
for estimation scale selection, but have limitations in terms of remote expert sessions.
153
“Solon” and “Consensus-2” DSS, appended with special add-ons for aggregation of
estimates, provided in different scales, and for defining of an optimal resource distri-
bution (i.e., the one that maximizes the degree of strategic goal achievement), allow us
to cover the whole spectrum of problems from strategic planning cycle only when used
in combination. And still, certain aspects require improvement.
The review allows a DM, an expert session organizer, or a knowledge engineer, to
select DS tools, which are most suitable for specific problems the expert session is in-
tended to solve.
3 Some other aspects and directions of DS tools improvement
We feel, that there are some other aspects of DS process, which remain beyond the
scope of our overview, but still need to be taken into consideration during improvement
of existing and development of new DS tools. Let us briefly list these aspects.
1) Defining the relative competence of experts during group expert sessions. In order
to define the relative competence of experts in each of the issues under consideration,
we should be able to take several components into account: objective component, mu-
tual estimate, and self-estimate of expert group members, as well as the quality of in-
formation, provided by the expert during a specific session [40].
2) Ordinal-data-based decision-making techniques – for cases when experts can pro-
vide only ordinal comparisons of alternatives [41,42].
3) Consideration of peculiarities of expert data collection [7], such as human psycho-
physiological constraints [43], requirement of keeping the estimates within one order
of magnitude, defining of the necessary number of pair-wise comparisons to be per-
formed (if estimates are provided in the form of pair-wise comparisons) [44], oppor-
tunity to change the sequence of alternative presentation to the expert for comparison
[45].
4) Opportunity to organize feedback with experts in order to improve the quality (con-
sistency, compatibility, completeness, and detail) of expert information [46].
5) Opportunity to consider not only expert but also open-data-based information in the
DSS [47]. This information is directly related to the level of development of natural
language processing (NLP) tools, used for DS.
6) Opportunity to analyze situation development scenarios under changing impact of
this or that factor. Scenario analysis can be based on sensitivity analysis principles.
4 Conclusions
The paper presents a brief overview and comparative analysis of available DS tools
from the standpoint of their applicability to strategic planning problem-solving in
weakly structured subject domains.
It has been shown, that not a single automated DSS, featured in the review, can solve
all the range of problems, emerging in the process of strategic planning based on expert
and other information. We have suggested a set of constructive requirements to modern
DS tools. We have obtained a set of recommendations on selection of software tools
154
for solving DS problems in the process of strategic planning, as well as on improvement
of existing (and development of new) DSS.
In spite of increasing volumes of available open data, that can be used for DS, the
relevance of tasks, associated with obtaining and processing of expert information, and,
especially, expert knowledge, is growing. Consequently, the problem of development
and improvement of DS tools, allowing to obtain and process both expert and other
information most thoroughly and without distortions, also remains relevant.
The most promising directions of further research on the paper’s topic include im-
provement of DSS tools’ features, related to organization of remote expert sessions and
incorporation of data on subject domain from all available sources.
References
1. Kadenko, S.: Prospects and Potential of Expert Decision-making Support Techniques Im-
plementation in Information Security Area. In CEUR Workshop Proceedings. Vol-1813, pp.
8-14 (2016).
2. Saaty T.: Decision Making with Dependence and Feedback: The analytic Network Process.
RWS Publicaitons, Pittsburgh (1996).
3. Totsenko, V. Methods and Systems of Decision-making Support. Algorithmic Aspect (in
Russian). Kyiv, Naukova Dumka (2002).
4. Tsyganok V.V., Kadenko S.V., Andriichuk O.V. Using Different Pair-wise Comparison
Scales for Developing Industrial Strategies. Int. J. Management and Decision Making. Vol.
14, No 3. – pp. 224–250 (2015).
5. Balagura I., Kadenko S., Andriichuk O., Gorbov I. Defining Potential Academic Expert
Groups based on Joint Authorship Networks Using Decision Support Tools. Selected Papers
of the XIX International Scientific and Practical Conference on Information Technologies
and Security (ITS 2019) Kyiv 2019 – pp. 222-233 (2019).
6. Tsyganok, V., Kadenko, S., Andriichuk, O., Roik, P.: Usage of multicriteria decision-mak-
ing support arsenal for strategic planning in environmental protection sphere. Journal of
Multi-criteria Decision Analysis, 24 (5-6), 227-238 (2017).
7. Dodonov, A., Lande, D., Tsyganok, V., Andriichuk, O., Kadenko, S., Graivoronskaya, A.:
Information Operations Recognition. From Nonlinear Analysis to Decision-Making. Lam-
bert Academic Publishing. (2019).
8. Zgurovsky, M. Scenario Analysis as a System Foresight Methodology (in Ukrainian). Sys-
tem Research and Information Tecnologies, i.1, pp. 7-38 (2002).
9. Podberyozkin, A. Probable scenario of international situation development after 2021 (in
Russian). M., MGIMO-UNiversitet (2015).
10. Applications and Theory of Analytic Hierarchy Process. Decision Making for Strategic De-
cisions. De Felice, F. (Ed.). IntechOpen, (2016).
11. Power, D. J. Decision support systems: concepts and resources for managers. Westport,
Conn., Quorum Books (2002).
12. Super Decisions. Available online at: http://www.superdecisions.com/. Accessed
30.11.2020.
13. Expert Choice Desktop: Powerful Performance for Organizational Decision-Making. Avail-
able online at: http://expertchoice.com/products-services/expert-choice-desktop. Accessed
30.11.2020.
155
14. ISAHP proceedings: Available online at: http://www.isahp.org/proceedings/. Accessed
30.11.2020.
15. Mareschal B.,Brans J.-P. "PROMETHEE Methods", Ch 5 in: Figueira, J., Greco, S., and
Ehrgott, M., eds., Multiple Criteria Decision Analysis: State of the Art Surveys Series / New
York: Springer, (2005).
16. Decision-making Support System “1000minds” Available online at:
http://www.1000minds.com/. Accessed 30.11.2020.
17. Hansen P.,Ombler F. A new method for scoring multi-attribute value models using pairwise
rankings of alternatives. Journal of Multi-Criteria Decision Analysis. 15, pp. 87-107 (2008).
18. Hovanov N.,Yudaeva M., Hovanov K.Multicriteria estimation of probabilities on basis of
expert non-numeric, non-exact and non-complete knowledge / European Journal of Opera-
tional Research. Vol. 195, Issue 3. – P. 857-863 (2009).
19. Decision-making Support System “Analytica” Available online at: https://lumina.com/prod-
ucts/free101/. Accessed 30.11.2020.
20. Decision-making Support System “D-sight” Available online at: http://www.d-sight.com/.
Accessed 30.11.2020.
21. Decision-making Support System “MakeItRational” Available online at: http://makeitra-
tional.com/. Accessed 30.11.2020.
22. Decision-making Support System “MindDecider” Available online at: http://www.mindde-
cider.com/. Accessed 30.11.2020.
23. Decision-making Support System “Logical Decisions” Available online at: http://www.log-
icaldecisions.com/. Accessed 30.11.2020.
24. Decision-making Support System “Tree Age Pro” Available online at:
http://www.treeage.com. Accessed 30.11.2020.
25. Decision-making Support System “RFP (Rational Focal Point)” Available online at:
https://www.teamblue.unicomsi.com/products/focal-point/. Accessed 30.11.2020.
26. Decision-making Support System “Very Good Choice” Available online at: https://mvlsoft-
very-good-choice-add-in-for-exce.software.informer.com/download/. Accessed
30.11.2020.
27. Roy B. Classement et choix en présence de points de vue multiples (la méthode ELECTRE).
La Revue d'Informatique et de Recherche Opérationnelle (RIRO). N8. – P. 57-75. (1968).
28. Информационно-аналитическая система “ОЦЕНКА И ВЫБОР” (Information and ana-
lytical system “Estimation and Choice”) Available online at: http://www.deol.ru/users/De-
cisionSupporter/projects/iasctc.html. Accessed 30.11.2020.
29. Demo Project – Compare car models 2001-2012 / ESTIMATION & CHOICE Available
online at: http://decisionsupporter.com/Projects.asp. Accessed 30.11.2020.
30. Mikony, S. V. “SVIR’” – a system for selection and ranking. (in Russian) Proceedings of an
international congress “Artificial intelligence in the XXI century”. Divnomorskoye 3–
8.09.2001. Vol. 1, pp. 500-507 (2001).
31. Mikony, S. V., Burakov, D. P., Sorokina, M. I. Implementation of ergonomics and intelli-
gence principles in SVIR’ system. Software products and systems, #3, 28-32 (2002).
32. Solon-3 DSS. Available online at: http://www.dss-lab.org.ua/Solon-3.htm. Accessed
30.11.2020.
33. System "Consesnsus-2" for distributed collecting of expert information for DSS. Available
online at: http://www.dss-lab.org.ua/Consensus-2.htm. Accessed 30.11.2020.
34. Medical Sapiens platform. Available online at: https://www.medicalsapiens.cl/. Accessed
30.11.2020.
35. Total Decision DSS. Available online at: https://vilenio.com/td_download.html. Accessed
30.11.2020.
156
36. Decision Mentor app. Available online at: https://www.decisionmentor.app/. Accessed
30.11.2020.
37. Best Decision Support Software | 2020 Reviews of the Most Popular Tools & Systems,
Available online at: https://www.capterra.com.au/directory/30544/decision-support/soft-
ware. Accessed 30.11.2020.
38. Weistroffer, H.R., Smith C.H., and Narula, S.C. Multiple criteria decision support software.
Ch 24 in: Figueira J., Greco S. and Ehrgott M. eds, Multiple Criteria Decision Analysis:
State of the Art Surveys Series / N.Y.: Springer, (2005).
39. Buckshaw D. Decision analysis software survey / OR/MS Today. Volume 37, #5. – P. 44-
53, (2010).
40. Kadenko, S.: Defining Relative Weights of Data Sources during Aggregation of Pair-wise
Comparisons. In CEUR Workshop Proceedings. Vol-2067, pp. 47-55 (2017).
41. Tsyganok, V., Kadenko, S. On sufficiency of the consistency level of group ordinal esti-
mates. Journal of Automation and Information Sciences. 42(8). P. 42–47 (2010).
42. Kadenko, S. Determination of Parameters of Criteria of" Tree" Type Hierarchy on the Basis
of Ordinal Estimates. Journal of Automation and Information Sciences. 40(8). P. 7-15
(2008).
43. Miller G. A. The Magical Number Seven, Plus or Minus Two. The Psychological Review.
63, 81-97, (1956).
44. Wedley, William C. Fewer Comparisons - Efficiency via Sufficient Redundancy. Proceed-
ings of the 10th International Symposium for the Analytic Hierarchy Process - Pittsburgh,
PA, July 30 - August 3, 2009. Available online at: http://www.isahp.org/2009Proceed-
ings/Final_Papers/94_Wedley_Fewer_Comparisons_REV_FIN.pdf. Accessed 30.11.2020.
45. Andriichuk, O., Tsyganok, V., Kadenko, S., and Porplenko, Y. Experimental Research of
Impact of Order of Pairwise Alternative Comparisons upon Credibility of Expert Session
Results. Proceedings of 2020 IEEE 2nd International Conference on System Analysis &
Intelligent Computing (SAIC). Kyiv, 5-9 Oct., 2020. DOI:
10.1109/SAIC51296.2020.9239126 (2020).
46. Tsyganok V., Kadenko S., Andriichuk O., and Roik P. Combinatorial Method for Aggrega-
tion of Incomplete Group Judgments. Proceedings of 2018 IEEE 1st International Confer-
ence on System Analysis & Intelligent Computing (SAIC). Kyiv, 08-12 Oct., 2018. P. 25–
30 (2018).
47. Tsyganok, V., Kadenko, S., and Andriichuk, O. Hybrid Decision Support Methodology
Based on Objective and Expert Data. 2020 IEEE 11th International Conference on Depend-
able Systems, Services and Technologies (DESSERT), Kyiv, Ukraine, 2020, pp. 265-271
(2020).