Multi agent paradigm for cognitive parameter based feature similarity for social influence

IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163
__________________________________________________________________________________________
Volume: 02 Issue: 04 | Apr-2013, Available @ http://www.ijret.org 687
MULTI-AGENT PARADIGM FOR COGNITIVE PARAMETER BASED
FEATURE SIMILARITY FOR SOCIAL INFLUENCE
Swati Basak1
, Bireshwar Dass Mazumdar2
, Chandra Bhan Patheriya3
1, 3
Dept. of Computer Engineering, SITM, Lucknow India, 2
SMS Varanasi India
swati.basak.bhu@gmail.com, bireshwardm@gmail.com, chandrabhan2006@gmail.com
Abstract
The ABM methodology is a favorable approach to model and analyze complex social phenomena that may involve non-linear
feedback loops. It has been applied successfully to model a number of social phenomena involving different social processes and
organizational structures. Availability of cheap computing power and rich software resources has made ABM a widely used and
hence more popular methodology. A modeler using ABM however have be careful about choosing the right amount of detail (less and
more both can be problematic) and validating (internal and external) the model. Interpreting and analyzing results is also an involved
task. In this paper, we have demonstrated how ABM can be applied to model and analyze the voting preference formation and
resultant voting decisions of individuals in a population. The model assumes a two party system. We designed three versions of the
simulation and observed the results for a large number of runs with different parameter variations. The results obtained present
interesting picture and resultant inferences.
--------------------------------------------------------------------***------------------------------------------------------------------------
1. INTRODUCTION
The development of social simulation over the past half
century can be grouped into three broad periods: macro
simulation, micro simulation and agent-based models.
Sociologists, particularly computational sociologists, first tried
macro simulations to model problems in supply-chain
management, inventory control in a warehouse, urban traffic,
and spread of epidemics, migration and demographic patterns.
Macro simulations consist of a set of differential equations
that take a holistic view of the system. However, taking the
complete system as the unit of analysis had inherent
limitations. Microsimulations focused on the use of
individuals as the unit of analysis but continued with macro-
level forecasting used in macrosimulations. Though
microsimulations modeled changes to each element of the
population but they did not permit individuals to directly
interact or to adapt. The main focus remained forecasting
macro effects of public policies that alter individual behaviors.
ABM, the third methodology, takes a pure bottom-up
approach and keeps the individual at the centre. It allows
modeling individual actors in the population and their
interactions with each other as well as with the environment.
This makes it a highly suitable technique for analysis of
emergent group behaviors resulting only from local
interactions of individuals. Besides, there are many other
advantages of this approach.
In Agent-based modeling (ABM) a system is modeled as a set
of autonomous agents, who can perceive the environment and
act on the basis of some behavioral rules. The agents represent
the actors in the system; environment represents the
surroundings including neighbouring agents; and the
behaviour rules model the interaction of the agent with other
agents as well as with the environment. ABM can be used to
model a number of phenomena in varied domains like Markets
& Economy, Organizations, World Wide Web and Social
Systems etc. Availability of fast and cheap computing power,
coupled with other advances in computational sciences has
paved the way for use of ABM as a preferred modeling and
simulation technique. Since last few years ABM has become
the frontrunner tool of the Sociologists and Psychologists who
try to model social cognitive parameter based behaviours,
particularly the behaviour of groups and societies. A large
number of researches are now being carried out using this
generative approach to model and analyze social phenomenon,
such as spread of rumors, diffusion of ideas, emergence of
cooperation, emergence of Conventions & social norms,
evolution and dynamics of spread of religions and cultures etc.
This paper presents our experimental work on using ABM for
understanding and analyzing cognitive parameter based voting
preferences of individuals. We have modeled the process and
factors shaping and affecting individual voting preferences for
a particular political party according to cognitive parameter
based. To keep the model simple and consequently analysis
more accurate, we have taken individual voting preferences as
„0‟ and „1‟ representing a two party political system. Every
individual‟s final voting decision is shaped and affected by a
number of factors, such as his own perception about parties,
his family views & beliefs, perception of his friends, and his
neighbourhood preferences. Our simulation models the role
and relative weights of these factors in shaping an individual‟s
voting preferences. We have also incorporated social influence
theory in our model which states that similar persons are more

__________________________________________________________________________________________
likely to interact, and as a result of frequent interactions they
are likely to become more similar.
2. ROLE OF AGENT BASED MODELING IN
SOCIAL INFLUENCE
Traditional modeling approaches to social influence systems
relied on equation-based models that operate with a macro
perspective. They operate on population attributes & their
distributions and lack the focus on individual‟s role. Equation-
based models, though useful for macro-scale behaviour
predictions, fail to model social influence systems (or
processes) that lack central coordination, systems that are very
complex in terms of their interdependencies and systems
which produce novel emergent behaviours in absence of a
clear understanding of the collective phenomenon. Axtell [1]
takes this argument a level further and suggests that there are
three distinct uses of Agent-based computational models of
social influence systems: (a) when numerical realizations can
be proposed and solved: agent models can be used as social
simulations; (b) when a mathematical model can be
formulated but can only be partially solved: agent based
models can be a useful tool of analysis; and (c) when
mathematical models are either apparently intractable or
provably insoluble: agent based modeling is perhaps the only
technique available for systematic analysis. Availability of fast
& cheap computing power along with rich easy-to-use
software environments also favours the use of ABM in Social
Sciences.
Although technically simple, ABM is conceptually deep.
Modeling a complex social process requires high conceptual
clarity and analytical ability. A key issue, therefore, is to
decide when to use ABM for modeling social influence
systems. An indicative list of situations when it is better to
think and model in terms of agents is: (a) when there is a
natural representation of actors as agents; (b) when the
interactions between the agents are complex, non-linear,
discontinuous, or discrete; (c) when agents exhibit complex
adaptive cognitive parameter based behaviours; (d) when the
population or topology of interactions is heterogeneous; and
(e) when agents have spatial component to their cognitive
parameter based behaviours & interactions. ABMs in social
sciences involve human agents, whose behaviours are often
complex, irrational and subjective. Therefore, one needs to
think carefully about the social phenomenon at hand before
going for ABM. Further, the model needs to be built at the
right level of description, with only the right amount of
details. Unnecessarily adding complexity to a model can make
the analysis difficult and the model useless.
ABM is currently being applied to model a variety of complex
social phenomenon where simple local interactions generate
emergent system-level behaviours. Some representative &
relevant work can be found in [1], [2], [3] & [4]. Macy &
Willer [3] group most of the work on Agent-based modeling
of collective behaviours in two categories: (a) models of
emergent structure which includes works on cultural
differentiation, homophilous clustering, idea diffusion,
convergent behaviours and norms; and (b) models of emergent
social order which include viability of trust, cooperation and
collective action in absence of global control. Goldstone &
Janssen [10] also identify three similar themes for Agent-
based computational models of collective behaviour namely:
(a) patterns and organizations which include settlement
patterns & segregation, human group behaviours and traffic
patterns; (b) social contagion which include spread of ideas,
fashions, cultures & religions; and (c) cooperation which
include evolution of cooperation, trust & reputations and
social norms & conventions.
3. SOCIAL INFLUENCE OF COGNITIVE
PARAMETERS
The Social influence assumes that individuals (or agents) often
imitate cognitive parameter based good behaviours and
cultures through their interactions with other individuals by
using coordination and cooperation (cognitive parameters). Its
dynamics depends on familiarity of the interacting individuals,
density of the neighbourhood, and popularity and spatial
proximity of the other individuals. The model of social
influence developed in two ways: (a) it explicitly takes into
account that the effect of one cultural feature depends on the
presence or absence of other cultural features and behaviours;
and (b) it takes into account that similar individuals are more
likely to influence each other than dissimilar individuals.
Interesting models of social influence have been proposed by
[4], [6], [7], [8], [9] etc. Axelrod [6] and [9], in his social
influence model of dissemination of culture [6], [9] focused on
factors of local influence (tendency for people who interact to
become more similar) and homophile (tendency to interact
more frequently with similar agents). The more agents
interact, the more similar they become; and the more similar
they become, the more likely they are to interact. Axelrod
expected convergence and homogeneity as the outcome but
simulation shown that despite the strong converging pressure,
stable regions of diversity persisted.
Axelrod [6] and [9] basic model included sites arrayed on a
grid. These sites are the basic actors of the model. Each site
can interact only with its immediate neighbours. Agents who
are similar to each other interact with each other and become
more similar. Axelrod‟s model captures culture as a group of
features with multiple trait values per features (cognitive
parameters). However, the emphasis is not on the content of a
specific culture but on the way in which culture is likely to
emerge and spread. The simulations with varying parameters
regarding grid size, number of features and number of traits
per feature resulted in polarization, despite the only
mechanism for change being convergence towards a
neighbour. We have used the basic idea of Axelrod‟s culture
model to represent friends of an individual [7]. An

__________________________________________________________________________________________
individual‟s friend is believed to have cognitive parameter
based behavioural feature set (cognitive parameter set) similar
to the individual. The feature set similarity represents their
like mindedness of the individuals on various issues and hence
possibility of having similar voting preferences (example of
model) or at least being affected by each other‟s preferences.
An individual while deciding about his vote (example model)
is thus likely to be affected by his interactions with friends.
Take an example of cognitive social influence model of
voting; an individual makes up his voting preference through a
complex process involving non-linear feedback loop of
interactions with a number of entities. Individuals interact with
frequently with their family members and relatively less
frequently with other persons belonging to their social and
religious groups. An individual‟s final voting value is no
doubt a function of these cognitive factors. However, what is
more important to understand is that what is the extent of this
effect and the by what underlying process an individual finally
makes up his voting preference. This model can be solved by
Greedy Approach and Case Base Reasoning (CBR) method,
some weighted values are assigned to the cognitive parameters
to decide the final voting preference of an individual; the CBR
method will be discussed in next section.
4. VOTING MODEL WITH FEATURE
SIMILARITY
In this Voting Model initially setup is design in such a manner
that all the individuals in grid or a particular area are assigning
with randomly votes and randomly assign feature set which
will decide final vote of individuals after dynamic interaction
among neighbour agents.
Model Setup
A population of agents is created on a wrap-around agent grid
of size 40 X 40. The key entities in the model are represented
by patches; each patch of the grid represents an agent. The
patches represent stationary landscape on the grid. In this
version agent hold a set of features. Each feature is a variable
that can adopt a binary value (1 or 0). Initially, agents adopt
randomly chosen a set of features. If feature size is n than
there will be 2n different type of feature set. Every agent has
one of this 2n feature set.
Model Parameters
In grid, we have placed agents at each coordinate. Every
individual has certain attributes that characterize it. Most
interestingly, it has a cognitive parameter based behavioural
feature set of size “n”, which can be a 1 to 10 bit string. Every
value in the string is a binary value (0 or 1) and represents a
particular feature. The feature set values of two agents is used
to decide whether they are sufficiently likeminded or not. Two
individuals having will be similar if their feature set string
contains a minimum number of similar feature values (greater
than a similarity threshold). An individual is more likely to be
influenced by those who are similar to him. In addition to the
feature set, every agent also has a voting preference value (1
or 0), initialized randomly. Before deciding his final voting
preference, an agent interacts with various agents in its
neighbourhood. According to social influence theory, it is
more likely to interact and be affected by the agents similar to
it. Interactions between similar agents make them more
similar. However, rather than changing the feature set values
of agents, we have changed their voting preference values so
that similar agents are more likely to have similar voting
preferences. In fact, the model picks up an agent and samples
its neighbourhood. Suppose n out of total m agents in the
neighbourhood are similar to the agent. Then the voting
preference of the agent changes to the majority preference of
the similar agents. In case of tie the agent retains its voting
preference. We obtained multiple runs of the model by
varying (i) the feature size, (ii) feature threshold size and (iii)
initial voting preference value distribution.
5. RESULTS
In these model multiples Runs are taken by varying feature
size, threshold size and initial voting preference value
distribution. The feature size “n” has been varied from 1 to 10
with feature similarity threshold being varied from 1 to n. Fig.
1 to Fig. 7 presents some initial and final snapshots of voting
model with feature similarity.
1(a) 1(b)
Fig. 1: A snapshot of the run of feature set based simulation.
The feature set size is 7 and similarity threshold is 1. The
initial values of green = 840 and blue = 841 (Fig. 1(a)),
transform to final values of green = 856 and blue = 825 and
stability time tick =10 (Fig. 1(b))
Table 1 shows few more of sampling runs by taking the above
data which describes the stability of model; time to stability
increases with decreasing the threshold size. It also shows if
threshold size nearer to feature size there will be less
polarization

__________________________________________________________________________________________
Table 1: Sample Runs of Feature Based Voting Model
CONCLUSIONS
The ABM methodology is a favorable approach to model and
analyze complex social phenomena that may involve non-
linear feedback loops. It has been applied successfully to
model a number of social phenomena involving different
social processes and organizational structures. Availability of
cheap computing power and rich software resources has made
ABM a widely used and hence more popular methodology. A
modeler using ABM however have be careful about choosing
the right amount of detail (less and more both can be
problematic) and validating (internal and external) the model.
Interpreting and analyzing results is also an involved task. In
this paper, we have demonstrated how ABM can be applied to
model and analyze the voting preference formation and
resultant voting decisions of individuals in a population. The
model assumes a two party system. We designed three
versions of the simulation and observed the results for a large
number of runs with different parameter variations. The results
obtained present interesting picture and resultant inferences.
The initial voting preference distribution in the population,
role of friends (likeminded individuals), affect of family
members and neighbours; all produce interesting results to
analyze. Social influence theory incorporated into voting
dynamics plays an important role. Individuals‟ voting
decisions are affected more by those who are similar to them.
However, if the similarity threshold is relatively lower, the
model results in local voting preference polarizations. When
the model is based only on majority, the initial distribution of
voting preferences becomes the key deciding factor. Both
these results taken together present similarity to the actual
voting patterns observed in closely knit societies and
segregations such as far off villages and tribal groups. The role
of implicit favorable waves introduced by mass media can also
be understood by the results of the second version of the
simulation. When the weights of factors playing role in
shaping an individual‟s voting preference are varied, the
results demonstrate that high weight to individual vote is able
to sustain variations in voting preferences. When the own-
weight is reduced to give more favour on the basis of
cognitive values to the family and neighbours, polarizing
tendency gains comes in and slight amount of local
polarization is seen in even in case of almost uniform initial
voting preference distribution. Voting behaviour and patterns
of relatively informed and well-educated persons conform to
this trend. The experimental simulations attempt to model an
interesting social cognitive phenomenon and present
interesting results that need to be further analyzed to draw
relevant inferences, analogies and explanations.
REFERENCES
[1]. J.M. Epstein, “Generative Social Sciences: Studies in
Agent-based Computational Modeling”, Princeton University
Press, (2007).
[2] J.M. Epstein & R. Axtell, “Growing Artificial Societies:
Socail science from bottom-up”, Cambridge, MA: MIT Press,
(1996).
[3]. M.W. Macy & R. Willer, “From factors to actors:
computational sociology and agent based modeling”, Annual
Review of Sociology, Vol. 28, pp. 143-166, (2002).
[4]. N. Gilbert & K. Troitzsch, “Simulation for the Social
Scientist”, Buckingham, U.K.: Open Univ. Press, (1999).
[5].R. Axelrod, “The Complexity of cooperation”, Princeton
N.J., Princeton University Press, (1997).
No. of
Runs
Fe
at
ur
e
Si
ze
Th
res
hol
d
Siz
e
No. of Green
Vote
No. of
Blue Vote
Tim
e to
Sta
ble
Initi
al
Final Initi
al
Fin
al
Run-1 7 1 840 856 841 825 10
Run-2 7 2 840 960 841 721 7
Run-3 7 3 840 860 841 821 7
Run-4 7 4 840 834 841 847 3
Run-5 7 5 840 838 841 843 2
Run-6 7 6 840 841 841 840 2
Run-7 7 7 840 840 841 841 1
Run-8 5 3 169 104 151
2
157
7
4
Run-9 5 3 505 417 117
6
126
4
3
Run-
10
5 3 673 610 100
8
107
1
4
Run-
11
3 2 169 144 151
2
153
7
3
Run-
12
3 2 505 458 117
6
122
3
2

__________________________________________________________________________________________
[6]. R. Axelrod, “The Dissemination of Culture: A model with
local convergence and global polarization”, Journal of
Conflict Resolution, Vol. 41, pp.203-226, (1997).
[7]. R. Axelrod & L. Tesfatison, “A guide for newcomers to
agent based modeling in the social sciences”, In the Handbook
of Computational Economics Vol. 2: Agent Based
Computational Economics, (2005).
[8].R. Axtell, “Why agents? On the varied motivations for
agent computing in the social sciences, working paper 17”,
Centre on Social and Economic Dynamics, Brookings
Institution, Washington D.C., (2000).
[9]. R. Axtell, R. Axelrod, J.M. Epstein & M.D. Cohen,
“Aligning simulation models: a case study and results”,
Computational and Mathematical Organization Theory, Vol.1,
No.2, pp.123-41, (1996).
[10]. R.L. Goldstone & M.A. Janssen, “Computational models
of collective behaviour”, Trends in Cognitive Sciences, Vol.9,
No.9, (2005).

Multi agent paradigm for cognitive parameter based feature similarity for social influence

More Related Content

What's hot (19)

Viewers also liked (17)

Similar to Multi agent paradigm for cognitive parameter based feature similarity for social influence (20)

More from eSAT Journals (20)

Recently uploaded (20)

Multi agent paradigm for cognitive parameter based feature similarity for social influence