Application of IFS Code for Conflict Resolution

Modeling and Agent Based Simulation

Shpend Ismaili1 and Stefka Fidanova2

1
University of Tetovo, Macedonia
shpend.ismaili@unite.edu.mk
2
Institute of Information and Communication Technology
Bulgarian Academy of Science, Bulgaria
stefka@parallel.bas.bg

Abstract. Very actual now days is modeling and simulation of conflict

situations. The aim is their peaceful prevention and conflict resolution.
Different mathematical approaches are used. One of them is applica-
tion of multi-agent systems. The multiagent system consists of software
agents, which are autonomous subject working together. In our applica-
tion the agents represents police officers and civilians in case of protest.
In this work we propose application of intuitionistic fuzzy sets for repre-
sentation of the movement and interaction of the agents. The intuitionis-
tic fuzzy sets are extension of the fuzzy sets including level of uncertainty.
Thus the model becomes more realistic. The conflict situation can cause
mess and casualties. Crowd simulation is very difficult and important
research topic.

1 Introduction
Crowd control and conflict situation prevention and elimination are very impor-
tant when some protest appear. Some of the factors are difficult to be predicted,
even when is expected that the crowd is well managed. Various approaches are
applied for researching crowd behavior, fuzzy-theory-based method [6], bandit
strategy [2], cellular automata [9], crowd motion simulation [8].
Multi aget systems is one of the methods which is applied on modeling crowd
behavior. Multi-agent system consists of various kind of agents and environ-
ments. The behavior of an agent can be changed by the interaction with other
agents and with environment. The agents can be passive or active and can react
in different manner according the situation [14].
The Intuitiounistic Fuzzy Sets (IFS) are proposed by Atanassov [1] as gen-
eralization of the fuzzy sets, incorporating degree of hesitation margin. Later,
several extensions and developments are proposed by Vassilev et al. in [11–13,
18, 19]. Some applications of IFS are done in [15]. Description of a problem or
process by IFS becomes more realistic. In this paper IFS are used for modeling
and simulation of conflicts, based on agents.
The rest of the paper is organized as follows. In Section 2, we give short
description of IFS. In Section 3 the problem is defined. In Section 4 the IFS are
2 Sh. Ismaili, S. Fidanova

applied for modeling and simulation of conflicts based on agents. Section 5 gives
and example for intutionistci fuzzy estimation of conflict situation. At the end
we give some conclusions.

2 Short Description of the IFS

The fuzzy sets [22] has meaningful application in many fields of study, but in
some real applications is possible the sum of the degree of membership and non-
membership to be less than 1. Sometimes there is an hesitation degree. Thus
Atanassov [1] proposed generalization of fuzzy sets called intuitionistic fuzzy sets,
which incorporate the degree of hesitation. So IFS has three components, degree
of membership, degree of non-membership and degree of hesitation (uncertainty)
and the sum of the values of the three components is equal to 1. The notion
of IFS is useful in many application. In [16] is shown the usefulnes of IFS in
problem with linguistic variables. IFS are tool for a more consistent reasoning
under imprecisely defined facts [17]. In [3] is given an IFS application in medical
diagnosis.
To have an idea about IFS first we will give an illustrative example. A good
example for IFS is weather forecast. Let the weather forecast for today is 60%
cloudy 30% sunny and 10% partially sunny partially cloudy. Thus there is 10%
uncertainty. The strong definition of the IFS is as follows:

Definition 1. Let E is a set. An IFS A in E is defined as an object of the

following form:
A = {hx, µA (x), νA (x)i|x ∈ E},
where µA : E → [0, 1], νA : E → [0, 1], define the degree of membership and
the degree of non-membership of the element x ∈ E, respectively, and for every
x ∈ E:
0 ≤ µa (x) + νA (x) ≤ 1.
The value of πA (x) = 1 − µA (x) − νA (x) is called the degree of uncertainty
of the element x ∈ E to the IFS A.

There are operations and relations over IFS. Most important are inclusion,
intersection, equality, negation, conjunction, disconjunction, product, scalar mul-
tiplication etc. [21] .

3 Problem Formulation

A conflict is a situation where minimum two persons, strive to achieve their

goals. It is a dynamic process. In this paper we try to understand the human
behavior and its variations according the situation. We try to represent different
groups in a conflict with a help of IFS and will simulate collective behavior. We
focus on development of multi agent system to learn the behavior caused by the
interaction between the agents. Some computer models are exist for concrete
 Title Suppressed Due to Excessive Length 3

protests: model of trade protest [7]; the model of violence in London [4]; model
of revolution [10].
We create more individuals which interact between them, to model civil vi-
olence. The structure consists of individuals, environment and empirical rules.
Our software agents model polis officers and civilians. Accurate modeling of
their attributes is crucial to the description which is as much as possible closer
to human life and behavior in situations of unrest. Peaceful civilians are neutral
participant, but they can react to external or internal stimulus. Police officers
retain the order by the insertion of the activists in jail and through strategies
that choose depends on the success of the management and control of violence.
The police officers perform two tasks in a direct way: active arrest protesters
and move in space.
Civilians are much more complex individuals, than the police officers. Civilian
agent decides whether to be active or not. Typical of civilian agents is communi-
cation. The civilian agents can change from active to passive and from passive to
active. The functioning of the system depends of the empirical rules. Empirical
rules guide the interactions of agents and ensure the functioning of the system.

4 Intuitionitic Fuzzy Sets for Conflict Resolution

We will apply IFS to represent the movement of the individuals, taking in to
account the rules in the network, which represents the space where the agents
are move. We specify three types of agents: peaceful civilians; police officers and
active civilians.

Fig. 1. Possible movements of agents

Every one of the agents has rules of movement. There are 8 possible direc-
tions. We will describe the agents movement in a 10 × 10 grid, which has 100
patches. It means that the all possible movement for one agent are 99, n = 99,
and the possible directions are m = 8, thus there are 91 impossible patches
for one step movement. In the case when there is not other agent around and
the current agent is not to the border of the grid (region), where some of the
movements are impossible, than:
 
8 91
V (agent) = ,
99 99
If the agent is on the border of the grid or in some of the corners or there is
other agent in on of the neighbor patches, than the possible directions are less.
4 Sh. Ismaili, S. Fidanova

For example let an agent is in the corner of the grid, then the possible directions
are 3 and:
 
3 91
V (agent) = ,
99 99
If an agent is on the grid border, then the possible directions are 5 and:
 
5 91
V (agent) = ,
99 99
We will introduce the terms active and peaceful and level of discontent (NAI)
and threshold (Athreshold) for danger [5], where NAI=Rev-N; Rev-tendency to
revolt and N is a net risk(the risk of imprisonment). There are several kind of
interactions between the agents:
– The police officer agent (P ) approaches the active civilian agent to arrest
him;
– The active civilian agent (A) try to stay fare from the police officers;
– The peaceful civilian agent (M ), his movement depends of value of NAI-
AThreshold. It is is great than 0, then it will become active civilian. If it
is less than 0 continue to be peaceful and will circumnavigate the active
civilians.
When we apply IFS for concrete cases will be concrete situations, according
the conditions which we examine. Every movement can be estimated as useful,
harmful or indifferent. The main three types of agents will be defined according
the three main situation, on which they corresponds.
– agents police officers ap with f1p , f2p , . . . , fapp ;
– agents active civilian aA with f1A , f2A , . . . , faAA ;
– agents peaceful civilian am with f1m , f2m , . . . , famm .
Let φ(f x , t) is the number of the possible movements of the agent f x at the
moment t, where x ∈ {A, p, m}. Let φu (f x , t), φh (f x , t) and φi (f x , t) are the
number of the possible movements in the network at the time t, thus:

φu (f x , t) + φh (f x , t) + φi (f x , t) = φ(f x , t)
Pax
φu (f x , t)
µx (t) = Pi=1
ax x
(1)
i=1 φ(f , t)
Pax
φh (f x , t)
νx (t) = Pi=1
ax x
(2)
i=1 φ(f , t)
Pax
φi (f x , t)
πx (t) = Pi=1
ax x
(3)
i=1 φ(f , t)
Hence µx (t) + νx (t) + πx (t) = 1 and the pair (µx (t), νx (t)) is a intuitionistic
fuzzy estimation of the network, where are the agents in the moment t.
 Title Suppressed Due to Excessive Length 5

5 Example

Let our network is a 10 × 10 grid, which consists of 100 patches and let there
are 8 agent of type P (polis officers), 12 agents of type A (active) and 8 agents
of type M (peaceful), which are located as it is shown on Figure 2.

a b c d e f g h i j

Fig. 2. Agents positioning

There are several possible movements for every type of agents. For the 8
agents of type P the possible movements in our example are:

– b0 → a0, a1, c0, c1;

– b1 → a0, a2, c0, c1, c2;
– a3 → a2, b3, b4;
– a4 → a5, b3, b4, b5;
– b6 → a5, a6, a7, b5, b7, c5, c6, c7;
– d7 → c6, c7, c8, d6, d8, e6, e7, e8;
– f 8 → e7, e8, e9, f 9, g7, g8, g9;
– j5 → i4, i5, i6, j4, j6;

The number of possible useful, harmful and indifferent movements at the

moment t is shown in Table 1. According Formulas ??,??, ??, for the values of
µx (t), νx (t) and πx (t) we receive:
Pax
φu (f x ,t) φub0 +φub1 +φua3 +φua4 +φub6 +φud7 +φuf 8 +φuj5 14
µx (t) = Pi=1
ax
φ(f x ,t)
= φb0 +φb1 +φa3 +φa4 +φb6 +φd7 +φf 8 +φj5 = 46 = 0.304
i=1
6 Sh. Ismaili, S. Fidanova

Table 1. Possible movements in different situations

agent position b0 b1 a3 a4 b6 d7 f8 j5
possible movements φ=4 φ=6 φ=3 φ=4 φ=8 φ=8 φ = 85 phi = 5
useful movements φu = 1 φu = 2 φu = 1 φu = 2 φu = 1 φu = 3 φu = 1 φu = 3
harmful movements φh = 2 φh = 3 φh = 1 φh = 1 φh = 6 φh = 3 φh = 3 φh = 1
indifferent movements φi = 1 φi = 1 φi = 1 φi = 1 φi = 1 φi = 2 φi = 4 φi = 2

6 Conclusion

Acknowledgment
Work presented here is partially supported by the Bulgarian National Scien-
tific Fund under the grants DFNI-I02/20 Efficient Parallel Algorithms for Large
Scale Computational Problems” and DN02/10 “New Instruments for Knowledge
Discovery from Data, and their Modeling”.

References
1. Atanassov K., Intuitionistic duzzy sets, Fuzzy Sets and Systems 20, 1986, 87-96.
2. Chen H., Rahwan I., Cebrian M., Bandit strategies in social search: the case of the
DARPA red balloon challenge, EPJ Data Science, Vol. 5(1), 2016, p.20.
3. De S.K., Biswas R., Roy A.R., An application of intuitionistic fuzzy sets in medical
diagnostic, Fuzzy sets and systems 117(2), 2001, 209-213.
4. Davies, T. P., Fry, H. M., Wilson, A. G., Bishop, S. R.: A mathematical model of
the London riots and their policing, Scientific Reports, vol. 3, n 1303, February
2013.
5. Epstein, J. M., Modeling civil violence: An agent-based computational approach,
Proceedings of the National Academy of Sciences of the United States of America,
vol. 99, 2002, pp. 72437250.
6. Fu L., Song W., Lo S., A fuzzy-theory-based method for studying the effect of in-
formation transmission on nonlinear crowd dispersion dynamics, Communications
in Nonlinear Science and Numerical Simulation, Vol. 42, 2017, pp. 682-698.
7. Kim, J. W., Hanneman, R. A.: A Computational Model of Worker Protest, Journal
of Artificial Societies and Social Simulation, vol. 14(3), 2011
8. Li D., Yuan L., Hu Y., Zhang, X., Large-scale crowd motion simulation based on
potential energy field, J. of Huazhong University of Science and Technology, Vol.
44(6), 2016, pp. 117-122.
9. Lubas R., Was J., Porzycki J., Cellular Automata as the basis of effective and
realistic agent-based models of crowd behavior, Journal of Supercomputing, Vol.
72(6), 2016, pp. 2170-2196.
10. Makowsky, M. D., Rubin, J.: An Agent-Based Model of Centralized Institutions,
Social Network Technology, and Revolution, Working Paper 2011-05, Towson Uni-
versity Department of Economics, 2011.
11. Marinov, E., Atanassov, K., Vassilev, P., Su, J., Directed intuitionistic fuzzy neigh-
bourhoods, 2016, 2016 IEEE 8th International Conference on Intelligent Systems,
IS 2016 - Proceedings, 544-549.
 Title Suppressed Due to Excessive Length 7

12. Marinov, E., Tsvetkov, R., Vassilev, P., Intuitionistic fuzzy inclusion indicator of
intuitionistic fuzzy sets, 2016, Studies in Fuzziness and Soft Computing, 332, 41-53.
13. Marinov, E., Vassilev, P., Atanassov, K., On separability of intuitionistic fuzzy
sets, 2016, Advances in Intelligent Systems and Computing, 401, 111-123.
14. Salamon T., Design of Agent-Based Models: Developing Computer Simulations
for a Better Understanding of Social Processes, ISBN 978-80-904661-1-1, Bruckner
Publishing, 2011.
15. Sotirov, S., Sotirova, E., Melin, P., Castilo, O., Atanassov, K. (2016). Modular
Neural Network Preprocessing Procedure with Intuitionistic Fuzzy InterCriteria
Analysis Method. In Flexible Query Answering Systems 2015, Springer Interna-
tional Publishing,175-186
16. Szmidt E., Kacprzyk J., Intuitionistic fuzzy sets in some medical applications, Note
on IFS 7(4), 2001, 58-64.
17. Szmidt E., Kacprzyk J., Medical diagnostic reasoning using a similarity measure
for intuitionistic fuzzy sets, Note on IFS 10(4), 2004, 61-69.
18. Vassilev, P., A note on the intuitionistic fuzzy set operator F,, 2004, 2004 2nd
International IEEE Conference ’Intelligent Systems’ - Proceedings, 3, 18-20.
19. Vassilev, P., Intuitionistic fuzzy sets generated by archimedean metrics and ultra-
metrics, 2017, Studies in Computational Intelligence, 657, 339-378.
20. Vassilev, P., Intuitionistic fuzzy sets with membership and non-membership func-
tions of exponential type, 2012, IS’2012 - 2012 6th IEEE International Conference
Intelligent Systems, Proceedings, 145-149.
21. Wang W., Liu X., Some Operations over Atanassov’s Intuitionistic Fuzzy Sets
Based on Einstein T-norm and T-conorm, J. of Uncertainty, Fuzziness and
Knowledge-Based Systems, Vol 21(2), 2013, 263-276.
22. Zadeh L.A., Fuzzy sets, Information and Control 8, 1965, 338-353.