Content uploaded by Motohisa Funabashi
Author content
All content in this area was uploaded by Motohisa Funabashi
Content may be subject to copyright.
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
Proposal for a Social-MRC: Social Consensus Formation Support System
Concerning IT Risk Countermeasures
Ryoichi Sasaki†, Shoko Sugimoto%, Hiroshi Yajima†,
Hidetaka Masuda†, Hiroshi Yoshiura＃, Masaki Samejima＄
† Tokyo Denki University, ({sasaki, yajima, masuda}@im.dendai.ac.jp)
% AdIn Research Inc., (sugimoto@adin.co.jp)
＃
The University of Electro-Communications (yoshiura@hc.uec.ac.jp)
＄
Osaka University, (samejima@ist.osaka-u.ac.jp)
doi: 10.4156/ijipm.vol2. issue2.6
Abstract
The problem of multiple risks is that a measure to reduce one risk can increase other risks. To
address this problem, a system to support social consensus formation is needed to mitigate the IT risks
such as those involved in information filtering to protect children. In cases in which the number of
people necessary for consensus formation is low, such as forming a consensus within an organization,
the Multiple Risk Communicator (MRC) developed by the authors offers a possible solution to this
problem. However, the MRC cannot be applied to problems in which the number of stakeholders
exceeds several thousand, and thus an innovative solution was considered necessary. Accordingly, the
authors modified and expanded the MRC and developed the concept “Social-MRC” to
comprehensively support risk communication on two levels: communication among opinion leaders
and communication allowing the participation of ordinary stakeholders. By introducing this two-level
approach, we are able to achieve consensus formation even in cases in which there are several
thousand or more stakeholders.
Keywords: -IT Risk, Risk Communication, Risk Management
1. Introduction
As illustrated by the example of how using bioethanol as a solution to an energy problem led to a
food problem, it is clear that we live in an era in which efforts to reduce one risk may increase another
risk, and so multiple risks must be taken into account [1][8]. This situation has led to an increase in the
number of problems requiring social consensus formation, such as those related to information filtering
to protect children. Accordingly, there is increasing need for a system that supports consensus
formation on such issues while taking into account multiple risks.
In cases in which the number of people necessary for consensus formation is low, such as within an
organization, the possibility of finding solutions is improved by use of the Multiple Risk
Communicator (MRC) developed by the authors [2]-[4]. However, the MRC cannot be applied to
problems for which the number of people involved exceeds several thousand, and thus an innovative
solution is necessary.
To achieve this solution, the authors proposed the concept “Social-MRC”, a social consensus
formation support system that comprehensively supports two-level multiple-risk communication
comprising communication among opinion leaders and participatory communication by ordinary
stakeholders. At the first level (communication among opinion leaders), not only quantitative but also
qualitative functions were added to the previously developed MRC to produce “MRC-Studio”. At the
second level (participation in discussion by ordinary stakeholders), MRC-Plaza was developed to
reflect the opinions of ordinary stakeholders in the deliberations of opinion leaders by using a video
sharing service, such as Ustream, to provide live broadcasts of deliberations by opinion leaders to
ordinary stakeholders and also showing the stakeholders the output of MRC-Studio being viewed by
the opinion leaders.
- 48 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
In accordance with the dual-process theory of risk psychology [7], we introduced both (1) a
systematic route, by which stakeholders having high motivation and ability with respect to the matter
in question make rational judgments, and (2) a heuristic route by which other stakeholders can express
their support or opposition to the judgments of the opinion leaders. For the introduction of opinions in
(1), we decided to use a microblog system, such as Twitter, to gather information, and graph theory and
natural language processing technology etc. to develop a method of (semi-)automatically analyzing
opinions of high importance or interest and effectively reflecting them in the deliberations of opinion
leaders.
To obtain social consensus, several prior examples exist, such as (a) consensus formation methods
based on the trans-science of gene therapy and genetically modified farm produce [5], and (b)
development of a risk communication support system and sharing information for reaching social
consensus on the geological disposal of high-level radioactive waste [6]. As security evaluation
methods for IT systems, the methods described in [11][12] were proposed. However, no systems
supporting social consensus formation solve the problem of conflicting risks by considering not only
major risks but also factors such as derivative risks, convenience, and cost. The great advantage of
Social-MRC is it can analyze the multiple risks and eliminate substantial delays in the implementation
of necessary social countermeasures that involve complex discussions.
2. MRC
2.1. MRC Development Requirements
The following is a summary of the development requirements of the MRC previously developed by
Figure 1. MRC development requirements and background
the authors. (Refer to the left side of Fig. 1).
Requirement 1: Many risks (e.g., security risks and privacy risks) exist. Accordingly, a means of
avoiding conflict among risks is necessary.
Requirement 2: It is difficult to achieve objectives by applying only single measures. Accordingly, a
system that seeks the optimal combination of measures is necessary.
Requirement 3: Many stakeholders (e.g., managers, customers, and employees) exist. Accordingly,
a risk communication method that can obtain consensus among many stakeholders is necessary.
We decided that the MRC devised to fulfill these requirements should formulate combinatorial
optimization problems having many risks and costs as constraints, to satisfy Requirements 1 and 2, and
solve these problems using an optimization engine while changing the values of the parameters and
constraints until stakeholder consensus is obtained, to satisfy Requirement 3.
To incorporate these functions into the MRC, the MRC was configured with the following (see the
right-hand side of Fig. 1): an input and output function for specialists, a computing function, a
stakeholder support function, an overall control function, a database function, and a negotiation
Requirement 2: It is difficult to
achieve objectives by applying only
single measures. Accordingly, a
system that seeks the optimal
combination of measures is necessary.
Requirement 1: Many risks (e.g.,
security risks and privacy risks) exist.
Accordingly, a means of avoiding
conflict among risks is necessary.
Requirement 3: Many stakeholders
(e.g., managers, customers, and
employees) exist. Accordingly, a
communication method that can obtain
consensus among many stakeholders is
necessary.
＜MRC＞
＜Requirements＞
Input and output
function for specialist
Optimization
engine
Simulator
Negotiatio n infrastruc ture
Overall co ntrol functio n
Stakeholder
support function
Specialist
Stakeholders
Facilitator
Computing function
Database
function
- 49 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
infrastructure. The users of this MRC are a specialist in MRC, stakeholders, and a facilitator who
mediates among these parties.
The MRC program was implemented using Java and PHP in a Windows XP environment.
The total number of coding steps in the original MRC was approximately 10,000.
2.2. MRC Application Procedure
The MRC application procedure is as follows (see Fig. 2).
The specialist of the MRC and the sponsor discuss and decide the problem to be solved. The
specialist makes advance preparations for inputting data into the MRC program (Fig. 2, (1)–(7)).
Then, the specialist inputs the data into the MRC program (Fig. 2, (8)). As previously
mentioned, although the specialist decides the constraint variables (Fig. 2, (4)), the stakeholders
can propose to change the constraint values, and the specialist also inputs these values into the
MRC program (Fig. 2, (9)). The MRC program performs the optimization calculation, and the
results are displayed (Fig. 2, (10)).
The specialist gathers together the stakeholders and shows them the data obtained from the
MRC program, and the stakeholders engage in risk communication to form a consensus on the
combination of countermeasures by expressing opinions such as “other countermeasures can be
Figure 2. MRC application procedure
considered” or “the constraint values are different” while viewing the results (Fig. 2, (11)).
The facilitator performs overall management and operation of the MRC. When the
stakeholders use the MRC program to engage in Web-based risk communication, the facilitator
manages the stakeholder login IDs and passwords and promote discussion among the
stakeholders by using a tool, such as a bulletin board in the MRC program. When gathering the
stakeholders together to engage in the risk communication, the facilitator contacts the
stakeholders and provides guidance or coordination to ensure that thorough discussion occurs
within the risk communication.
Please refer to Ref. [2] for more details on the application procedure.
2.3 MRC Application Examples
Previously, the MRC was applied to personal information leakage problems, internal control
problems, and other matters [2]-[4]. For example, the Setagaya-ku government office applied
the MRC to solve a personal information leakage issue for all junior high schools in Setagaya-
ku and is currently preparing to implement the countermeasures recommended and agreed upon.
In this case, the authors can confirm its basic effectiveness.
The MRC can effectively handle the following: (1) cases in which a few representative
stakeholders gather together and seek consensus formation, or (2) cases of simulated consensus
(1) Decision on the problem to be solved
(2) Understanding of the problem
(3) Decision on the stakeholders
(e.g., citizens)
(4) Decision on the objective function
and constraints
(5) Fault tree analysis (FTA),
event tree analysis (ETA), etc.
＜Advance Preparations＞
Specialists
(8) Data input into the MRC
program
(9) Decision on constraint
values
(10) Optimization calculation
(11) Results display and
consensus formation
Satisfied ? END
yes
no
<Use of the MRC Program>
Facilitator
(7) Decision on parameters
Stakeholders
(6) Decision on proposed measures
- 50 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
formation through interactive role-playing in groups to simulate the stakeholders in companies,
local government bodies, or other organizations. The MRC cannot be applied to problems of
social consensus formation among several thousand or more stakeholders, and so an innovative
solution became necessary.
3. The Social-MRC: Social Consensus Formation Support System
3.1. Overall Concept of the Social-MRC
To fulfill the above requirements, we developed the concept of the Social-MRC, a social consensus
formation system that comprehensively supports two-level multiple-risk communication:
communication among opinion leaders and participatory communication by ordinary stakeholders (See
Fig.3). At the first level (communication among opinion leaders), necessary functions based on the
previously developed MRC were added and the new system was named MRC-Studio. At the second
level (participation in discussion by ordinary stakeholders), MRC-Studio was developed to reflect the
opinions of ordinary stakeholders in the deliberations of opinion leaders by using a microblog system,
such as Twitter, and a video sharing service, such as Ustream, to provide live broadcasts of
deliberations by opinion leaders to ordinary stakeholders and also showing ordinary stakeholders the
MRC-Studio output being viewed by the opinion leaders.
Figure 3. Overview of Social-MRC
Topics to which the Social-MRC can be applied include information filtering to protect children, the
use of surveillance cameras and its effect on privacy, and the introduction of a citizen identification
system. Possible situations in which the system is suitable include Web-based public hearings,
consensus meetings, television discussion programs, and other situations in which it is desirable to
have a thorough discussion and rapid consensus formation. The Social-MRC should be useful for
engaging in two-way information exchanges that incorporate the opinions of stakeholders, such as
citizens, as an alternative to one-sided exchanges, to decide policies while confirming the potential
impact of changing certain parameters on the overall situation.
Social-MRC
Opinion
leaders
Facilitator
Support server
MRC
specialist
Ordinary
stakeholders
< Level One >
MRC-Studio
(1) Support for
consensus formation
among opinion
leaders
(2) Support for
reflecting the
opinions of ordinary
stakeholders
< Level Two >
MRC-Plaza
Live
broadcast
of meeting
(1) Live broadcast of
meeting or MRC-Studio
output display
(2) Provision of
information to the
facilitator through
automatic analysis of
ordinary stakeholder
opinions
(Newly developed)
(Expansion of the
MRC)
Problems to
be solved
Use scenes Web-based public hearings, consensus meetings, government
program reviews, television discussion programs
Information filtering to protect children, introduction of a citizen
identification system, installation of surveillance cameras
Internet
- 51 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
3.2. Social-MRC Operation Method
Figure 4 shows the system configuration when the abovementioned Social-MRC is used in a
consensus formation concerning IT risk countermeasures. The procedure is as follows.
Phase 1 Preparation for the Broadcast
(1) The sponsor decides in advance the problem to be solved and the opinion leaders.
(2) The specialist formulates the problem to be solved as a combined optimization problem, inputs
the parameter and constraint values into MRC-Studio, and seeks an optimal combination of measures
as an initial solution.
(3) The specialist shows the results to the opinion leaders, adds proposed measures, changes
parameter values, changes constraint values, and uses MRC-Studio to calculate the optimal
combination of the proposed measures for each opinion leader (see Fig. 5).
Phase 2 Selecting the Preferred Opinion Leader
(1) Each opinion leader expresses his or her preferred combination of proposed countermeasures
obtained by using MRC-Studio in an advance deliberation along with the following:
(a) Basic viewpoint
(b) Evaluation indexes that should be emphasized
(c) Stakeholders who should be considered
(d) Proposed countermeasures that should be completely rejected
(e) Expected results
Figure 4. Social-MRC system configuration
(2) This process is shown to the ordinary stakeholders using images captured by video cameras
through MRC-Plaza. The ordinary stakeholders select their preferred leader.
(3) As a result, the opinion leaders having the most supporters are selected by MRC-Plaza.
Phase 3 Forming the Consensus among Opinion Leaders
(1) The results of the selection are relayed to the facilitator and opinion leaders via MRC-Plaza.
The subsequent discussion is focused on the combination of proposals of the selected opinion leader.
(2) Each opinion leader points out problems regarding the combinations of proposed
countermeasures or makes observations, such as differences in the parameter values and constraints.
4
Router
Ustream
server
Twitter
server
Overall display screen image (for
example, Conference feed, MRC
output, Stakeholder response )
MRC-
Studio
server
MRC-Plaza
server
Conference room
Camera
Ordinary
stakeholders
Internet
＜Social-MRC＞
Chairperson
Overall
director
Syste
matic
route
Heuri
-stic
route
Writing
down of
opinions
Supporter
selection
Director in
charge of MRC-
Plaza
Opinion
leaders
＜MRC-Plaza＞
＜MRC-Studio＞
<MRC output,
Conference feed>
<Opinions>
<Selection>
Director
in charge
of MRC-
Studio
MRC
specialist
<Opinions>
- 52 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
In response, the MRC specialist calculates the optimal combination of proposed countermeasures
using the MRC-Studio optimization engine, and displays the results on the display screen.
Figure 5. Example of optimization results for each opinion leader
(3) This process is made known to the ordinary stakeholders. The ordinary stakeholders indicate
the opinion leaders they support and input their own opinions. MRC-Plaza carries out the following:
indicates which opinion leaders’ opinions have the highest support, semi-automatically analyzes the
important opinions, and conveys the results to the facilitator and opinion leaders.
(4) The processes in (2) and (3) of Phase 3 are repeated until a provisional consensus is formed or
until a deadline is reached.
Phase 4 Stakeholders’ Voting on the Provisional Optimal Solution Obtained by the Opinion
Leaders
(1) Stakeholders answer questions such as “Do you agree with the provisional optimal solution by
the opinion leaders?”
(2) If the majority says “YES”, the risk communication process is finished. If the majority says
“NO,” the process goes back to (2) in Phase 3 and the consensus formation is continued.
Phase 5 Arrangements after Broadcasting
(1) The results of the consensus formation are linked to specific countermeasures. Various
approaches can be considered, depending on the sponsor. In the case of additionally introducing
public opinion, such as the opinions of customers and employees, into the consensus formation within
an organization, the sponsor decides and implements the countermeasures. In the case of information
filtering countermeasures approved by the government, the countermeasures to be taken are legislative
that is, the countermeasures become laws.
(2) The specialist or facilitator analyzes the Social-MRC application process and records the
process for use in a future application.
(3) In the case that a deadline is reached without a formed consensus, the sponsor plans the next
conference.
3.3. Issues to Be Resolved in MRC-Studio
Generally, it is not easy for users of the MRC or MRC-Studio to provide quantitative values for the
parameters of the objective function and/or constraint functions. Therefore, we need specialists and
opinion leaders to easily decide the values as follows.
(1) Ensure that the initial modeling and the basis for the parameter value calculation by the MRC
specialist are organized in an easy-to-understand way and facilitate the addition of modifications and
modeling by the opinion leaders. Although the original MRC has a basic function for this, make the
system easier to use, including the aspects pertaining to expertise.
5
Combination of measures
3–5, 8, 10, and 14
Optimal value
Constraint values and
other values
Alice’s optimal
solution
Bob’s optimal solution
- 53 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
(2) In the performance of simulations and the calculation of optimal combinations using the
previous MRC, only one point value could be applied to the parameters (for example, costs or the
probability of an obstacle occurring). However, there are cases in which it is difficult to apply such
value, and there are many cases in which the application of a range of values is desirable. Accordingly,
enable the application of the parameter values by using ranges or distributions and make possible the
following after the conversion of qualitative values into random variables.
(a) Develop a qualitative-quantitative combined simulator to perform simulations such that it is
possible to calculate and display the values of the objective function and the constraint formulas in the
form of probability distributions [9].
(b) Furthermore, enable the calculation of the optimal combination of proposed measures even in
cases in which each parameter has a probability distribution. It is theoretically possible to solve such
problems by formulating them as chance constraint programming problems [10], an optimization
technique. For example, consider the case in which, in each constraint formula, the total cost parameter
Ci (i = 1, 2, ..., n) has a probability distribution, the mean is C’i (i = 1, 2, ..., n), and the standard
deviation is σi (i = 1,2, ...,n). These constraints can be applied as follows:
n
Prob(ΣCi•Xi ≤ Ct) ≥ 95%, (1)
i=1
where Ct is the total cost constraint and Xi is a 0/1 variable that indicates the adoption or non-
adoption of proposed measure i, with Xi = 0 indicating non-adoption and xi = 1 indicating adoption.
The probability on the left-hand side of Eq. 1 is the probability that the condition in parentheses holds;
here, 95% signifies that the probability must be 95% or higher.
Assume that the distribution of Ci is a normal distribution or, even if it is not, that the value of n is
sufficiently large, so that the formula can be modified using the central limit theorem as follows:
n n
ΣC’i•X i≤ Ct － αΣδi•Xi. (2)
i=1 i=1
When the above conditions hold, it is known that α is approximately 1.6. And, in this case, it is
possible to calculate the optimal solution as before. However, since ordinarily the distribution of each
parameter is not a normal distribution and n may not be sufficiently large, it is necessary to verify by
trial and error whether Eq. 2 holds.
Accordingly, after setting the value for α on the assumption of a normal distribution and calculating
the optimal combination of proposed measures, an experiment should be conducted with the approach
shown in Fig.6, which involves conducting a qualitative-quantitative combined simulation and
observing the constraint value distributions. If the probability that holds is near the target (e.g., 95%),
then that is the solution. If the probability is less than the target, make minor adjustments to the value
of α so that the result approaches the target. How specifically α should be adjusted is a task for future
research.
(3) Make it possible to rapidly and appropriately ascertain the opinions of ordinary stakeholders.
Do this as shown in (1) above.
(4) Establish an easy-to-understand display method for the overall display screen in Fig. 4.
Refine it through experimentation and other means.
- 54 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
Figure 6. How to deal with both quantitative data and qualitative data
3.4. Issues to Be Resolved in MRC-Plaza
The following is a discussion of the issues and the measures proposed to resolve these issues.
Issue 1—Make it possible to reflect the wishes of many people (thousands) in the consensus
formation.
Implement the following measures to achieve this.
(1) As previously mentioned, it is not the case that all participants access a single MRC. As
shown in Fig. 4, the proposed consensus formation process is divided into the opinion leader
conference stage (handled by MRC-Studio) and the ordinary stakeholder opinion gathering stage
(handled by MRC-Plaza). A Social-MRC (social consensus formation MRC) system is being
developed to realize two-level comprehensive support.
(2) To support participation by many ordinary stakeholders, this development makes use of
existing Internet-based systems. At this time, as shown in Fig. 4, we are considering using the Ustream
video sharing service for the live broadcast of conferences and using the microblog Twitter for
stakeholders to express their opinions. This is likely to facilitate the provision of information to several
thousand stakeholders and enable them to express their opinions.
(3) According to the dual-process theory of risk psychology, two types of people exist: rational
judges who can properly express their own opinions on the basis of the systematic route and people
who can only indicate the people whose opinions they agree with on the basis of the heuristic route [7].
Because both types of opinions are valuable, we decided to use Twitter to enable the rational judges to
freely express their opinions and to enable the other participants to select which opinion they prefer.
Issue 2—Make it possible to display to ordinary stakeholders in an easy-to-understand way the
discussions of the opinion leaders and the consensus formation process.
Fig. 6 shows a simulation of the screen that ordinary stakeholders will view over the Internet. Here,
we plan a framework for (1) displaying the feed of the opinion leader conference shown via Ustream,
(2) displaying the MRC-Studio output screen, (3) enabling people to freely enter their opinions using a
Twitter microblog function, and (4) enabling people to indicate whose opinions they support.
In the MRC, it is not easy to quantitatively apply a discrete parameter value.
In cases in which it is difficult tolerate ambiguity and apply values as
illustrated below.
Value
1.00.5
Value
0.5 1.0
For example, C1For example, C3
77
QQCS
Parameter Value Distribution
Constraints
X1=X3=1
C1 C３
Ct
９５％？
Σci・Xi
i=1
ｎ
Satisfied?
END
YES
MRC-Studio
Optimization engine
Calculation of C’i, the mean of Ci, and δi, the standard deviation
Convert to ΣC’iXi ≤Ct－αΣδiXi
and solve the combination optimization
problem (X1 = X3 = 1, X2 = 0).
Prob(∑Ci・Xi≦Ct)≧９５％ （Xi =0 or 1）
Set αassuming a normal distribution.
Adjust the value of α.
Formulation as a chance constraint programming problem
No
QQCS：Qualitative-quantitative combined simulator
- 55 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
Figure 7. The MRC-Plaza screen image (for ordinary stakeholders)
Arrange to show visitors to the Social-MRC website information such as “Conference name,”
“Conference purpose,” “Conference date and time,” “Ustream URL used,” and “Twitter hash ID used”
and to display screens such as that shown in Fig. 7 when visitors follow the instructions on the screen.
Issue 3—Make it possible to rapidly ascertain the opinions of large numbers of people and indicate
those opinions to the facilitator and the opinion leaders.
(1) Show the facilitator and the opinion leaders the results of whose opinions selected by the
ordinary stakeholders in the form of bar graphs or other statistical methods.
(2) Since the number of written opinions may be very large, (semi-)automatically select the
opinions that are to be shown to the facilitator and the opinion leaders in accordance with the following
concepts.
(a) When the same opinion is expressed with high frequency, the opinion is worth showing to the
facilitator and the opinion leaders. To that end, it is necessary to develop a technology that uses natural
language processing technology to organize the written opinions according to patterns and to rank them
in the order of frequency.
(b) The opinions of people who have triggered changes in the opinion trend and the opinions of
others which increase substantially as a result are important. Make it possible to discover and display
the opinions of people who have triggered changes in the opinion trend by using graph logic to analyze
the opinions linked by the Twitter Reply function or Retweet function.
(c) The recording of agreement with the opinions of the opinion leaders who have the highest
support by using the heuristic route is important. It is necessary to consider how to reflect the support
rate of the opinion leaders in the display of opinions.
(d) Opinions on the constraint values and other matters necessary for input into the MRC and
recalculation are important. A technology that enables the automatic determination of opinions
concerning constraints from among the various opinions is necessary.
Modifications concerning these matters will be added through the analysis of actual data, as this
is an important technical issue. The results should be displayed in a way that makes it easy for the
facilitator and the opinion leaders to understand the opinion situation and the correlation among
opinions. The specific method for this display is a matter for future research.
4. Development of prototype program of MRC-Plaza
We developed the prototype program of MRC-Plaza using JAVA and open Web technologies such
as microblogs and a video sharing service. The prototype program of MRC-Plaza and the conventional
MRC program used as a prototype program of MRC-Studio were applied to the information filtering
(1) Video shown using Ustream
(Opinion leader conference feed) (2) MRC-Studio output results
(3) For opinion input
(4) Input screen for
supporter selection
- 56 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
issue to protect children as the first small-scale experiment. In this experiment, the number of opinion
leaders was two and the number of stakeholders was five.
Fig. 8 shows the display of the prototype program of MRC-Plaza to the stakeholders. The
stakeholders were able to watch the meeting conducted by opinion leaders by using the Ustream
function, so that the stakeholders could know the output of the MRC. Moreover, the stakeholders were
able to input their opinion using the Twitter function and select options using the original function.
Although this experiment was very limited, we were able to confirm that the prototype program of
Figure 8. Display of MRC-Plaza for Selecting the Preferred Opinion Leader
MRC-Plaza performed the expected functions.
The total evaluation of the Social-MRC is a future issue.
5. Future Plans
(1) Improve the Social-MRC prototype to perform application experiments on concrete problems.
We are considering application to consensus building on topics such as countermeasures against
personal information leakage, information filtering to protect children, and surveillance cameras.
(2) Develop the actual Social-MRC and apply it to actual problems. In actual application tests,
we intend to apply the Social-MRC to cases in which the number of ordinary stakeholders exceeds
1,000.
6. Conclusion
We have reported on a modification of the MRC, which the authors developed for use in consensus
formation within organizations, and on the development of the Social-MRC system for
comprehensively supporting two-level multiple-risk communication consisting of communication
among opinion leaders and ordinary stakeholder participatory communication.
We will improve the Social-MRC and apply it to the information filtering issue for children
in an environment where several thousands of actual stakeholders are participating.
6
Selecting the
preferred
opinion
leader
MRC-Studio
output
Opinions of stakeholders from Twitter
Broadcast by Ustream of the meeting
Selection with Twtpoll
- 57 -
Proposal for a Social-MRC: Social Consensus Formation Support System Concerning IT Risk Countermeasures
Ryoichi Sasaki, Shoko Sugimoto, Hiroshi Yajima,, Hidetaka Masuda, Hiroshi Yoshiura, Masaki Samejima
International Journal of Information Processing and Management. Volume 2, Number 2, April 2011
7. Acknowledgement
Research concerning the MRC was conducted within the Japan Science and the Technology Agency
Research Institute of Science and Technology for Society’s Information and Society program and the
Identification and Solving of Vulnerabilities in an Advanced Information Society program of planned
research in the information and technology area from 2002 to 2007 and within the SECOM Science
and Technology Foundation Securing the Information Society with the Total Security Architecture
Design Incorporating Security Fundamental Technologies project from 2007 to 2009.
8. Reference
[1] Ryoichi Sasaki , “How to deal with IT risk” Iwanami, 2008
[2] Ryoichi Sasaki, Yuu Hidaka, Takashi Moriya, Mituhiro Taniyama, Hiroshi Yajima, Kiyomi
Yaegashi, Yasumasa Kawashima, Hiroshi Yoshiura, ” Development and applications of a
multiple risk communicator” Sixth International Conference on RISK ANALYSIS 2008(in
Greece) 2008.5
[3] Mitsuhiro Taniyama, Yuu Hidaka, Masato Arai, Satoshi Kai, Hiromi Igawa, Hiroshi Yajima and
Ryoichi Sasaki, “Application of ‘Multiple Risk Communicator’ to the Personal Information
Leakage Problem”, Proceedings of world academy of science, engineering and technology,
volume 35, pp.285-290, 2008
[4] Takashi Moriya, Hiroyuki Chiba, Ryoichi Sasaki, “Proposal and Application of Evaluation Method
Considering various Risks and Stakeholders for the Internal Control “ Japan Society of Security
Management, Vol. 22, No.3 pp3-14, 2008
[5] Denji Kobayashi, “Who consider Science and Technology - Experiments Named Consensus
Meeting” Nagoya University Publishing, 2004
[6] http://de.gsec.keio.ac.jp/rcsystem/
[7] Kazuya Nakagaichi, “Secure but Uneasy”, Chikuya, 2008
[8] Bruce Schneier, ” Beyond Fear: Thinking Sensibly About Security in an Uncertain World”Springer,
2003
[9] Masaki Samejima, Masanori Akiyoshi, Ryoichi Sasaki, “Support for Social Consensus using on
qualitative and quantitative hybrid simulation ”Information Processing Society of Japan, 49th
CSEC Conference, May, 2010
[10] R.S.Gerfinkel et al., “Integer Programming”, Wiley and Sons, (1972)
[11] Yacov Y. Haimes, Stan Kaplan, James H. Lambert, “Risk Filtering, Ranking, and Management
Framework Using Hierarchical Holographic Modeling”, Risk Analysis, Vol. 22, No. 2 , pp383-
397, 2002.
[12] Rainer Boehme,” Security Metrics and Security Investment Models”, IWSEC 2010, Kobe, Japan,
November 2010
- 58 -
