No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Toward a Methodology for Agent-Based Data Mining and Visualization
Abstract
We explore the notion of agent-based data mining and visualization as a means for exploring large, multi-dimensional data sets. In Reynolds' classic flocking algorithm (1987), individuals move in a 2-dimensional space and emulate the behavior of a flock of birds (or "boids", as Reynolds refers to them). Each individual in the simulated flock exhibits specific behaviors that dictate how it moves and how it interacts with other boids in its "neighborhood". We are interested in using this approach as a way of visualizing large multi-dimensional data sets. In particular, we are focused on data sets in which records contain time-tagged information about people (e.g., a student in an educational data set or a patient in a medical records data set). We present a system in which individuals in the data set are represented as agents, or "data boids". The flocking exhibited by our boids is driven not by observation and emulation of creatures in nature, but rather by features inherent in the data set. The visualization quickly shows separation of data boids into clusters, where members are attracted to each other by common feature values.
... ABM simulation platforms should also provide direct visualization functionality to vividly visualize interaction between agents as well as visualize system development through the whole simulation [40]. Through efficient ABM visualization, the simulation platform can effectively convey the behavior of the model and helps the user to quickly understand the model's outputs [41,42]. ...
Significant computation challenges are emerging as agent-based modeling becomes more complicated and dynamically data-driven. In this context, parallel simulation is an attractive solution when dealing with massive data and computation requirements. Nearly all the available distributed simulation systems, however, do not support geospatial phenomena modeling, dynamic data injection, and real-time visualization. To tackle these problems, we propose a distributed dynamic-data driven simulation and analysis system (4D-SAS) specifically for massive spatial agent-based modeling to support real-time representation and analysis of geospatial phenomena. To accomplish large-scale geospatial problem-solving, the 4D-SAS system was spatially enabled to support geospatial model development and employs high-performance computing to improve simulation performance. It can automatically decompose simulation tasks and distribute them among computing nodes following two common schemes: order division or spatial decomposition. Moreover, it provides streaming channels and a storage database to incorporate dynamic data into simulation models; updating agent context in real-time. A new online visualization module was developed based on a GIS mapping library, SharpMap, for an animated display of model execution to help clients understand the model outputs efficiently. To evaluate the system’s efficiency and scalability, two different spatially explicitly agent-based models, an en-route choice model, and a forest fire propagation model, were created on 4D-SAS. Simulation results illustrate that 4D-SAS provides an efficient platform for dynamic data-driven geospatial modeling, e.g., both discrete multi-agent simulation and grid-based cellular automata, demonstrating efficient support for massive parallel simulation. The parallel efficiency of the two models is above 0.7 and remains nearly stable in our experiments.
... Data mining Agent which is the core of the system consist of some task-oriented functional module such as the feature analysis, relevant analysis, classification, and the estimation and deviation analysis and .Data mining Agent uses the methods of machine leaning and statistical analysis to find some useful model or knowledge to mine [3] . Knowledge warehouse is used for directing search or valuing whether the model of mining results is meaningful or not [4] . The function of model valuing Agent is analyzing the information taken on the basis of the final decision-making purpose, and separating the most valuable information, then, handing it in to the decision maker through the decision-making supporting tool. ...
A Multi-Agent based transmission fault diagnosis system is researched in this paper. Many data digging analysis methods are employed, combined with data warehouse, OLAP and Multi-Agent technology. An intelligent decision supporting system for monitoring transmission network data is built. Data digging method is used to intelligently analyze and process fault data in the data warehouse, and Agent technology is used to realize data collection, pretreatment, inquiry, knowledge Automatic extraction, mining and other functions, which makes the whole mining process intellectual and intelligent. It aids transmission management with decision-making, thus to make the monitoring and repair of power grid fault more timely and accurate.
We present a model of opinion dynamics in which agents adjust continuous opinions as a result of random binary encounters whenever their difference in opinion is below a given threshold. High thresholds yield convergence of opinions towards an average opinion, whereas low thresholds result in several opinion clusters: members of the ame cluster share the same opinion but are no longer influenced by members of other clusters.
We present a study of the properties of network of political discussions
on one of the most popular Polish Internet forums. This provides the opportunity to study the
computer mediated human interactions
in strongly bipolar environment. The comments of the participants
are found to be mostly disagreements, with strong percentage of invective
and provocative ones. Binary exchanges (quarrels) play significant role
in the network growth and topology. Statistical analysis shows that the growth of
the discussions depends on the degree of controversy of the subject and
the intensity of personal conflict between the participants. This is in
contrast to most previously studied social networks, for example networks of scientific
citations, where the nature of the links is much more positive and
based on similarity and collaboration rather than opposition and abuse.
The work discusses also
the implications of the findings for more general studies of consensus
formation, where our observations of increased conflict
contradict the usual assumptions that interactions
between people lead to averaging of opinions and agreement.
How do blogs cite and influence each other? How do such links evolve? Does the popularity of old blog posts drop exponentially with time? These are some of the questions that we address in this work. Our goal is to build a model that generates realistic cascades, so that it can help us with link prediction and outlier detection. Blogs (weblogs) have become an important medium of information because of their timely publication, ease of use, and wide availability. In fact, they often make headlines, by discussing and discovering evidence about political events and facts. Often blogs link to one an- other, creating a publicly available record of how infor- mation and influence spreads through an underlying so- cial network. Aggregating links from several blog posts creates a directed graph which we analyze to discover the patterns of information propagation in blogspace, and thereby understand the underlying social network. Here we report some surprising findings of the blog linking and information propagation structure, after we analyzed one of the largest available datasets, with 45,000 blogs and � 2.2 million blog-postings. Our anal- ysis also sheds light on how rumors, viruses, and ideas propagate over social and computer networks. We also present a simple model that mimics the spread of infor- mation on the blogosphere, and produces information cascades very similar to those found in real life.
An experiment with two computer-based sessions (complete data for 156 participants) examined selective exposure to attitude-consistent and counterattitudinal media messages. In the first session, participants indicated interest in politics and news, political attitudes, with four target issues embedded, along with attitude certainty and importance. Attitude accessibility data were derived from response latencies. In the second session, participants browsed an online opinion forum with eight texts about four issues, each with a pair of articles presenting opposing views. Selective exposure was unobtrusively recorded by software and coded as attitude-consistent and counterattitudinal based on individual participants' attitudes. Results show that attitude-consistent exposure dominated regardless of particular issue, with 36% more reading time. Higher habitual news use and attitude certainty both fostered attitude-consistent exposure. Selection of counterattitudinal articles was more likely among participants with greater interest in politics, conservative party preference, stronger party preference, more accessible attitudes, and higher attitude importance.
A new model for a random graph is proposed that can be constructed from empirical data and has some desirable properties compared to scale-free graphs (1, 2, 3) for certain applications. The newly proposed random graph maintains the same "small-world" properties (3, 4, 5) of the scale-free graph, while allowing mathematical modeling of the relationships that make up the random graph. E-mail communication data was collected on a group of 24 mid-career Army officers in a one-year graduate program (6) to validate necessary assumptions for this new class of random graphs. Statistical distributions on graph level measures are then approximated using Monte Carlo simulation and used to detect change in a graph over time.
We present an agent-based software environment for modeling and simulation of adaptive consumers responding to dynamic electricity pricing. It has been specially designed for scenarios involving household customers. Households can be modeled down to the layer of single appliances, even taking into account presence and price awareness of inhabitants. Modeled utilities can calculate prices from different factors using different methods. The focus of investigations conducted is the analysis of household load shifting potential under different tariffs and different negotiation strategies.
This paper illustrates the resampling-approach to a task of comparison of two renewal processes with application to solving one inventory theory problem. The probability of the shortage absence of inventory unit is estimated. The formula for the variance calculation of estimator of interest is presented. Obtained resampling-estimators are compared with classical ones. It is shown that considered approach could be a good alternative to classical one, taking bias, variance and mean square error of estimators into account. Numerical examples illustrate the efficiency of considered method.
We review the recent rapid progress in the statistical physics of evolving networks. Interest has focused mainly on the structural properties of complex networks in communications, biology, social sciences and economics. A number of giant artificial networks of this kind have recently been created, which opens a wide field for the study of their topology, evolution, and the complex processes which occur in them. Such networks possess a rich set of scaling properties. A number of them are scale-free and show striking resilience against random breakdowns. In spite of the large sizes of these networks, the distances between most of their vertices are short - a feature known as the "small-world" effect. We discuss how growing networks self-organize into scale-free structures, and investigate the role of the mechanism of preferential linking. We consider the topological and structural properties of evolving networks, and percolation and disease spread on these networks. We present a number of models demonstrating the main features of evolving networks and discuss current approaches for their simulation and analytical study. Applications of the general results to particular networks in nature are discussed. We demonstrate the generic connections of the network growth processes with the general problems of non-equilibrium physics, econophysics, evolutionary biology, and so on.
Cluster analysis involves the problem of optimal partitioning of a given set of entities into a pre-assigned number of mutually exclusive and exhaustive clusters. Here the problem is formulated in two different ways with the distance function (a) of minimizing the within groups sums of squares and (b) minimizing the maximum distance within groups. These lead to different kinds of linear and non-linear (0–1) integer programming problems. Computational difficulties are discussed and efficient algorithms are provided for some special cases.
A computer simulation modeled the change of attitudes in a population resulting from the interactive, reciprocal, and recursive operation of B. Latané's (see record
1982-01296-001) theory of social impact, which specifies principles underlying how individuals are affected by their social environment. Surprisingly, several macrolevel phenomena emerged from the simple operation of this microlevel theory, including an incomplete polarization of opinions reaching a stable equilibrium, with coherent minority subgroups managing to exist near the margins of the whole population. Computer simulations, neglected in group dynamics for 20 years, may, as in modern physics, help determine the extent to which group-level phenomena result from individual-level processes. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
The task of continuous online unsupervised learning of streaming data in complex dynamic environments under conditions of uncertainty is an NP-hard optimization problem for general metric spaces. This paper describes a computationally efficient adaptive multi-agent approach to continuous online clustering of streaming data, which is originally sensitive to environmental variations and provides a fast dynamic response with event-driven incremental improvement of optimization results, trading-off operating time and result quality. Experimental results demonstrate the strong performance of the implemented multi-agent learning system for continuous online optimization of both synthetic datasets and datasets from the RoboCup Soccer and Rescue domains.
One of the approaches used to improve the accuracy and relevancy in information retrieval is cluster analysis. Clustering methods determine relationships among text documents, and allow the determination of similar groups or clusters of documents. These methods are computationally expensive, thereby limiting their use to a relatively small set of documents. This paper describes a multi-agent system to cluster large data sets. This technique is then compared to hierarchical agglomerative clustering using a small set of text data. Results show that the agent-based approach can significantly reduce the time required to cluster large data sets.
Distributed data mining is originated from the need of mining over decentralised data sources. Data mining techniques involving
in such complex environment must encounter great dynamics due to changes in the system can affect the overall performance
of the system. Agent computing whose aim is to deal with complex systems has revealed opportunities to improve distributed
data mining systems in a number of ways. This paper surveys the integration of multi-agent system and distributed data mining,
also known as agent-based distributed data mining, in terms of significance, system overview, existing systems, and research
trends.
These sixty contributions from researchers in ethology, ecology, cybernetics, artificial intelligence, robotics, and related fields delve into the behaviors and underlying mechanisms that allow animals and, potentially, robots to adapt and survive in uncertain environments. They focus in particular on simulation models in order to help characterize and compare various organizational principles or architectures capable of inducing adaptive behavior in real or artificial animals.
Jean-Arcady Meyer is Director of Research at CNRS, Paris. Stewart W. Wilson is a Scientist at The Rowland Institute for Science, Cambridge, Massachusetts.
Bradford Books imprint
Due to the proliferation of online auctions, there is an increasing need to monitor and bid in multiple auctions in order to procure the best deal for the desired good. Against this background, this paper reports on the development of a heuristic decision making framework that an autonomous agent can exploit to tackle the problem of bidding across multiple auctions with varying protocols (including English, Dutch and Vickrey). The framework is flexible, configurable and enables the agent to adopt varying tactics and strategies that attempt to ensure the desired item is delivered in a manner consistent with the user's preferences. In this context, however, the best strategy for an agent to use is very much determined by the nature of the environmnet and by the user's preferences. Given this large space of possibilities, we employ a genetic algorithm to search (offline) for effective strategies in common classes of environment. The strategies that emerge from this evolution are then codified into the agent's reasoning behaviour so that it can select the most appropriate strategy to employ in its prevailing circumstances.
The predictive performance of Echo State neural networks were optimized for electrical load forecasting and compared to the results achieved by competitors in the worldwide Eunite Competition #1. The test data used were the actual results of the competition, attached to a specific region. A regular adaptation of an Echo State neural network was optimized by adapting the weights of the dynamic reservoir through Anti-Hebbian learning, and the weights from input and output neurons to the hidden neurons were optimized using the Metropolis algorithm. The results achieved with such an optimized Echo State neural network would gain a strong second place within the Eunite competition.
Conceptual clustering is an important way of summarizing and explaining data. However, the recent formulation of this paradigm has allowed little exploration of conceptual clustering as a means of improving performance. Furthermore, previous work in conceptual clustering has not explicitly dealt with constraints imposed by real world environments. This article presents COBWEB, a conceptual clustering system that organizes data so as to maximize inference ability. Additionally, COBWEB is incremental and computationally economical, and thus can be flexibly applied in a variety of domains.
Google's MapReduce programming model serves for processing large data sets in a massively parallel manner. We deliver the first rigorous description of the model including its advancement as Google's domain-specific language Sawzall. To this end, we reverse-engineer the seminal papers on MapReduce and Sawzall, and we capture our findings as an executable specification. We also identify and resolve some obscurities in the informal presentation given in the seminal papers. We use typed functional programming (specifically Haskell) as a tool for design recovery and executable specification. Our development comprises three components: (i) the basic program skeleton that underlies MapReduce computations; (ii) the opportunities for parallelism in executing MapReduce computations; (iii) the fundamental characteristics of Sawzall's aggregators as an advancement of the MapReduce approach. Our development does not formalize the more implementational aspects of an actual, distributed execution of MapReduce computations.
In agent-enhanced workflow, a community of intelligent, distributed and autonomous software agents is used to improve the management of business processes under the control of a workflow management system. These improvements are achieved by allowing the software agents to negotiate with each other to establish contracts that govern the distribution of work across a number of processing centres. Furthermore, the agents collaborate to perform real-time exception handling, and to co-ordinate the redistribution of work items to meet changing circumstances.
We study a model of opinion formation based on the theory of social impact and the concept of cellular automata. The case is considered when two strong agents influence the group: a strong leader and an external social impact acting uniformly on every individual. There are two basic stationary states of the system: cluster of the leader's adherents and unification of opinions. In the deterministic limit the variation of parameters like the leader's strength or external impact can change the size of the cluster or, when they reach some critical values, make the system jump to another phase. For a certain range of parameters multistability and hysteresis phenomena are observed. In the presence of noise (social temperature) the rapid changes can be regarded as the first-order phase transitions. When both agents are in a kind of balance, a second-order transition and critical behaviour can be observed. Another kind of noise-induced transitions are the inverses (flips) of the unified group opinion due to random flips of the leader's opinion. Analytical results obtained within a mean field approximation are well reproduced in computer simulations.
A failure model with damage accumulation is considered. Damages occur according to a Poisson process and they degenerate into failures in a random time. The rate of the Poisson process and the degeneration time distribution are unknown. Two sample populations are available: a sample of intervals between damages and a sample of degeneration times. The case of small samples is considered. The purpose is to estimate the expectation and the distribution of the number of damages and failures at time t. We consider the plug-in and resampling estimators of the above mentioned characteristics. The expectations and variances of the suggested estimators are investigated. The numerical examples show that the resampling estimator has some advantages.
Online auctions have become one of the fastest growing modes of online- commerce transactions. eBay has 94 million active members buying and selling goods at a staggering rate. These auctions are also producing large amounts of data that can be utilized to provide services to the buyers and sellers, market research, and product development. We collect historical auction data from eBay and use machine learning algorithms to predict end-prices of auction items. We describe the features used, and several formulations of the price prediction problem. Using the PDA category from eBay, we show that our algorithms are extremely accurate and can result in a useful set of services for buyers and sellers in online marketplaces.
Privacy is defined as the freedom from unauthorised intrusion. The availability of public records along with intelligent search engines and data mining tools allow easy access to useful information. They also serve as a haven for individuals with malicious intent. This paper proposes an approach that protects the privacy of individual records while retaining the information content. The techniques that have been proposed for privacy protection so far either provide insufficient privacy or too much useful information on account of privacy protection. This paper proposes an attack model to analyse the different types of privacy breaches, proposes a set of properties for good privacy protection, proposes a robust data protection technique, and compares the privacy and usability properties of the new technique with some of the existing techniques.
This paper is about binary choices with externalities. These are either-or situations,not choices of degree or quantity. An "externality" is present when you care about my choice or my choice affects yours. You may not care, but need to know-whether to pass on left or right when we meet. You may not need to know, but care---you will drive whether or not I drive--- but prefer that I keep off the road.You may both care and need to know. The literature of externalities has mostly to do with how much of a good or a bad should be produced,consumed,or allowed.HereI consider only the interdependence of choices to do or not to do, to join or not to join, to stay or to leave, to vote yes or no, to conform or not to conform to some agreement or rule or restriction. Players will accept hockey helmets (or not) by individual choice for several reasons. Chicago star Bobby Hull cites the simplest factor: "vanity." But many players honestly believe that helmets will cut their efficiency and put them at a disadvantage, and others fear the ridicule of opponents. The use of helmets will spread only through fear caused by injuries like Green's-or through a rule making them mandatory. . . One player summed up the feelings of many: "It's foolish not to wear a helmet. But I don't because the other guys don't. I know that' silly, but most of the players feel the same way. If the league made us do it, though, we'd all wear them and nobody would mind." Shortly after Teddy Green of the Bruins took a hockey stick in his brain, the player, Don Awrey, commented to a Newsweek(1969) reporter, "WhenI saw the way Teddy looked, it was an awful feeling . . . I'm going to start wearing a helmet now, and I don't care what anybody says." A voluntary helmet may be seen as cowardly,but nobody thinks a baseball player timid when he dons the batting helmet without which the league will not let him bat. Motorcycle helmets are not only worn regularly,but probably worn more gladly, in states that require them. When ever ascribed motives matter, the way a choice is organized or constrained will itself be a part of the "outcome" and affect the payoffs.
The work investigates the influence of leader on opinion formation in artificial networked societies. The strength of the social influence is assumed to be dictated by distance from one agent to another, as well as individual strengths of the agents. The leader is assumed to have much greater resources, which allows him to tune the way he influences the other agents. We study various strategies of using these resources to optimize the conditions needed to "convince" the whole society to leader's opinion. The flexibility of the model allows it to be used in studies of political, social and marketing activities and opinion formation.
We consider a class of simple estimators of the change-point m in a sequence of n independent random variables X1…,Xn satisfying E(Xi) = θ0 for i = 1,…,m and E(Xi) = θ0+δn for i = m +1,…n. (θ0 and δn are unknown). We obtain rates of consistency for the estimator, derive its limiting distribution and show that the bootstrap approximation is asymptotically valid. The results are illustrated by some simulations.
A model and simulation of the "Electric Enterprise" (taken in the broadest possible sense) have been developed. The model uses autonomous, adaptive agents to represent both the possible industrial components, and the corporate entities that own these components. An open access transmission application and real-time pricing has been implemented. Objectives are: 1) To develop a high-fidelity scenario-free modeling and optimization tool to use for gaining strategic insight into the operation of the deregulated power industry; 2) to show how networks of communicating and cooperating intelligent software agents can be used to adaptively manage complex distributed systems; 3) to investigate how collections of agents (agencies) can be used to buy and sell electricity and participate in the electronic marketplace; and ultimately to create self-optimizing and self-healing capabilities for the electric power grid and the interconnected critical infrastructures.
With more substantial funding from research organizations and industry, numerous large-scale applications, and recently developed technologies, the Semantic Web is quickly emerging as a well-recognized and important area of computer science. While Semantic Web technologies are still rapidly evolving, Foundations of Semantic Web Technologies focuses on the established foundations in this area that have become relatively stable over time. It thoroughly covers basic introductions and intuitions, technical details, and formal foundations. The book concentrates on Semantic Web technologies standardized by the World Wide Web Consortium: RDF and SPARQL enable data exchange and querying, RDFS and OWL provide expressive ontology modeling, and RIF supports rule-based modeling. The text also describes methods for specifying, querying, and reasoning with ontological information. In addition, it explores topics that are clearly beyond foundations, such as tools, applications, and engineering aspects. Written by highly respected researchers with a deep understanding of the material, this text centers on the formal specifications of the subject and supplies many pointers that are useful for employing Semantic Web technologies in practice. The book has an accompanying website with supplemental information.
Bell System Technical Journal, also pp. 623-656 (October)
We consider the problem of producing recommendations from collective user behavior while simultaneously providing guarantees of privacy for these users. Specifically, we consider the Netflix Prize data set, and its leading algorithms, adapted to the framework of differential privacy. Unlike prior privacy work concerned with cryptographically securing the computation of recommendations, differential privacy constrains a computation in a way that precludes any inference about the underlying records from its output. Such algorithms necessarily introduce uncertainty--i.e., noise--to computations, trading accuracy for privacy. We find that several of the leading approaches in the Netflix Prize competition can be adapted to provide differential privacy, without significantly degrading their accuracy. To adapt these algorithms, we explicitly factor them into two parts, an aggregation/learning phase that can be performed with differential privacy guarantees, and an individual recommendation phase that uses the learned correlations and an individual's data to provide personalized recommendations. The adaptations are non-trivial, and involve both careful analysis of the per-record sensitivity of the algorithms to calibrate noise, as well as new post-processing steps to mitigate the impact of this noise. We measure the empirical trade-off between accuracy and privacy in these adaptations, and find that we can provide non-trivial formal privacy guarantees while still outperforming the Cinematch baseline Netflix provides.
The using of least square support vector machine for on-line forecast has been gradually applied to the field on management science research. The traditional support vector machine algorithm contains inequality constraints, which requires solving quadratic programming problems so that the computing can be very complicated when there are a lot of training samples. In this paper, first of all, the least square support vector machine algorithm has been improved so as to solve the sparsity and time lag problems existing in traditional method, and then set up the LS-SVM on-line forecasting model of the least support value sample based on time factor eliminating, and input observed data on the network sale instances about one production into the model for testing. The results show that: the forecast and actual value are comfortably approximate, and can well indicate the trends of e-commerce sales forecast; the error between forecast and actual value from this method is smaller than the forecast error from common least square support vector machine method and BP neural network method.
Since its release in June 2005, Google Earth has been bringing satellite images of our planet into our homes, or at least to those homes with broadband connections. Computer users, excited by seeing their own houses from on high, or even their cars parked in the drive, have been raving about this impressive piece of software which can be downloaded to your PC (but not, as yet, your Mac) free of charge. After a weekend playing around with Google Earth, I can confirm that there is a whole range of potential applications for this software in teaching Earth science.
Electronic commerce, and in particular online auctions, have received an extreme surge of popularity in recent years. While auction theory has been studied for a long time from a game-theory perspective, the electronic implementation of the auction mechanism poses new and challenging research questions. In this work, we focus on the price formation process and its dynamics. We present a new source of rich auction data and introduce an innovative way of modelling and analyzing price dynamics. We represent auctions as functional objects by accommodating the special structure of bidding data. We then use functional data analysis to characterize different types of auctions. Our findings suggest that there are several types of dynamics even for auctions of comparable items. By profiling these differences with respect to features associated with the auction format, the seller and the winner we find new relationships between dynamics and auction settings, and we tie these findings to the existing literature on online auctions.
This chapter aims to demonstrate the potential of self-organization for data visualization purposes. At the same time, it also illustrates how self-organization probably is not the most efficient existing method to visually represent data, due to the required computation power and the considerable amount of efforts needed in managing the various parameters that control the emergent phenomena. As will be discussed in Sect. 13.5, self-organizing visualization suffers from inadequate calculation performance, comprehensibility difficulties, and the influence of multiple parameters on the emergent effects. However, self-organizing data visualization still forms a valuable alternative approach to the predetermined and fixed data mapping techniques that currently exist, as it specifically allows unexpected visual organizations to occur.
In this chapter, the principles of the self-organizing data visualization model are described and illustrated with three different case studies. The ‘information particle’, ‘information flocking’ and ‘cellular ant’ approaches each use a different visual metaphor, respectively based on particle animation, swarming and cellular automata principles. Each system demonstrates how meaningful properties contained within datasets can be made apparent by the simulation of visually emergent effects, using relatively simple self-organizing behavior rules. Based on these findings, Sect. 13.5 discusses the potential benefits and shortcomings of this concept.
This paper is devoted to apply agent based models
to real opinion interactions in a bulletin board system (BBS) community. By analyzing a real BBS
community, we reveal some empirical features of opinion interactions on the Web. Then, we propose a simple opinion model that
bears both
general characteristics of traditional opinion models, and some real
interacting rules on the Web. The model simulates a group of
agents representing Web users participating to a discussion. Simulation results show some dynamical regimes consistent with
empirical facts, and offer some possible explanations for the emergence of some real features. Our work implies the possibility
of building simple
agent based models to simulate computer-mediated interactions on the
Web.