Showing papers on "Network theory published in 2003"

Congestion and centrality in traffic flow on complex networks

Abstract: The central points of communication network flow have often been identified using graph theoretical centrality measures. In real networks, the state of traffic density arises from an interplay between the dynamics of the flow and the underlying network structure. In this work we investigate the relationship between centrality measures and the density of traffic for some simple particle hopping models on networks with emerging scale-free degree distributions. We also study how the speed of the dynamics are affected by the underlying network structure. Among other conclusions, we find that, even at low traffic densities, the dynamical measure of traffic density (the occupation ratio) has a non-trivial dependence on the static centrality (quantified by "betweenness centrality"), where non-central vertices get a comparatively large portion of the traffic.

Molecular networks: the top-down view.

Abstract: Network theory can give a useful overview of how a biological system works. But to make testable predictions, we need the details.

Communicating centrality in policy network drawings

Abstract: We introduce a network visualization technique that supports an analytical method applied in the social sciences. Policy network analysis is an approach to study policy making structures, processes, and outcomes, thereby concentrating on relations between policy actors. An important operational concept for the analysis of policy networks is the notion of centrality, i.e., the distinction of actors according to their importance in a relational structure. We integrate this measure in a layout model for networks by mapping structural to geometric centrality. Thus, centrality values and network data can be presented simultaneously and explored interactively.

An Empirical Analysis of the Network Structure of the Austrian Interbank Market 1

Abstract: We provide an empirical analysis of the network structure of the Austrian interbank market based on a unique data set of the Oesterreichische Nationalbank (OeNB). The analysis relies on the idea that an interbank market can be interpreted as a network where the banks form the nodes and the claims and liabilities between them define the edges of the network. This approach allows us to apply results from general network theory, which is widely applied in other scientific disciplines — mainly in physics. Specifically, we use different measures from this network theory to investigate the empirical network structure of the Austrian banking system. We focus on the question of how this structure affects the stability of the network (the banking system) with respect to the elimination of a node in the network (the default of a single bank). Regarding the network structure, we find that there are very few banks with many interbank linkages whereas there are many with only a few links. This feature of networks has been repeatedly found to be conducive to the robustness of the network against the random breakdown of links (the default of single institutions due to external shocks). In addition, the interbank network shows a community structure that exactly mirrors the regional and sectoral organization of the current Austrian banking system. Moreover, the banking network has typical structural features found in numerous other complex real world networks: a low clustering coefficient and a relatively short average shortest path length. These empirical findings are in marked contrast to network structures that have been assumed in the theoretical economic and econo-physics literature.

Economics of Standards in Information Networks

Abstract: 1 Introduction.- 1.1 Standards in Information Systems.- 1.2 Motivation and research questions.- 1.3 Structure of the thesis.- 1.4 Methodology and definitions.- 1.4.1 Methodology.- 1.4.2 Standard semantics.- 1.4.3 Empirical data from Fortune 1,000 study.- 2 Standardization problems.- 2.1 A brief history of standardization.- 2.1.1 From DNA to railways.- 2.1.2 From production side to demand side economies of scale.- 2.2 Network effect theory as theoretical foundation.- 2.2.1 Basics.- 2.2.1.1 General findings of network effect theory.- 2.2.1.2 Related literature.- 2.2.2 Path dependency, lock-in and multiple equilibria.- 2.2.3 Instability ("tippy networks").- 2.2.4 Excess inertia: The start-up problem.- 2.2.5 Excess momentum: Too much standardization.- 2.2.6 Internalizing network effects in sponsored networks.- 2.2.7 Pareto-inferior market results and monopolies.- 2.2.8 Laissez-faire vs. dirigisme.- 2.3 Reconsidering network effect theory.- 2.3.1 Network effects versus network externalities.- 2.3.2 Direct vs. indirect network effects.- 2.3.3 Inefficiency and market failure.- 2.3.4 Costs of network size.- 2.3.5 Homogeneous network effects.- 2.3.6 Confusion of centralized and decentralized decision making.- 2.3.7 Ambiguous empirical evidence.- 2.3.7.1 Keyboards: QWERTY vs. Dvorak's DSK.- 2.3.7.2 The battle for VCR standards: beta isn't always better.- 2.3.7.3 Nuclear power reactors.- 2.3.7.4 Automobiles: Gas vs. Gasoline.- 2.3.8 Normative Implications.- 2.3.9 Summary.- 3 A standardization framework.- 3.1 Modeling the standardization problem: A basic model.- 3.1.1 A standardization model for centrally coordinated networks.- 3.1.2 A standardization model for decentrally coordinated networks.- 3.2 Equilibria in standardization problems.- 3.2.1 Information.- 3.2.2 Basic games.- 3.2.3 Equilibria.- 3.2.4 Efficiency of equilibria.- 3.2.4.1 Pareto efficiency.- 3.2.4.2 Kaldor-Hicks efficiency.- 3.3 Standardization games.- 3.3.1 Standard standardization games.- 3.3.2 Static non-cooperative 2-player standardization games.- 3.3.3 Sequential non-cooperative 2-player standardization games.- 3.3.3.1 Complete, perfect information.- 3.3.3.2 Complete, imperfect information.- 3.3.3.3 Incomplete information.- 3.3.3.4 Bayes-Nash-Equilibria.- 3.3.4 Dynamic n-player standardization games.- 3.3.4.1 Incomplete, imperfect information.- 3.3.4.2 Incomplete, perfect information.- 4 A simulation model for the standardization problem.- 4.1 Simulation design.- 4.1.1 Decision functions.- 4.1.2 Simulation parameters and pattern.- 4.2 The basic standardization problem.- 4.2.1 The single-period basic standardization problem.- 4.2.1.1 The standardization gap.- 4.2.1.2 Stability of the standardization gap.- 4.2.2 The multi-period basic standardization problem.- 4.2.2.1 The standardization gap.- 4.2.2.2 Single standard implementation vs. continuous license costs.- 4.2.3 Sequential choice.- 4.2.4 Individual consequences.- 4.2.4.1 Individual gains and losses.- 4.2.4.2 Wrong decisions.- 4.2.5 Agent size and standardization time.- 4.2.5.1 Stationary state period and agent size.- 4.2.5.2 Stationary state period and network size.- 4.2.6 Variation of network structure.- 4.2.6.1 Errors in different topologies.- 4.2.6.2 Standardization costs and network density.- 4.2.6.3 Summary.- 4.2.7 Installed base effects.- 4.2.8 Summary of multi-period one standard problems.- 4.3 The extended standardization model.- 4.3.1 The standardization gap and the equilibrium process.- 4.3.1.1 Simultaneous decisions, single choice.- 4.3.1.2 An adapted Herfindahl index.- 4.3.1.3 Simultaneous choice with reversible decisions.- 4.3.2 Five diffusion patterns.- 4.3.2.1 No standardization.- 4.3.2.2 Mixed solution.- 4.3.2.3 Monopoly.- 4.3.2.4 Oligopoly.- 4.3.2.5 Dynamic equilibrium.- 4.3.2.6 Comparing the diffusion paths.- 4.3.2.7 Summary.- 4.3.3 Sequential choice.- 4.3.4 Individual consequences.- 4.3.5 Agent size.- 4.3.6 Variations of network structure.- 4.3.7 Installed base effects.- 4.4 A different decentralized anticipation calculus.- 4.5 Adapting the diffusion model.- 4.5.1 From standardization to diffusion.- 4.5.1.1 The diffusion model.- 4.5.1.2 An integration.- 4.5.1.3 Further research.- 5 Solution mechanisms and empirical data.- 5.1 Empirical data for the standardization problem.- 5.1.1 Standardization problems in corporate networks.- 5.1.2 The importance of network effects.- 5.1.3 Diffusion of EDI-Standards.- 5.1.4 EDI data: costs and benefits.- 5.1.4.1 Empirical EDI benefits in the literature.- 5.1.4.2 Empirical EDI costs in the literature.- 5.1.5 EDI data from large enterprises.- 5.1.6 EDI data from SMEs.- 5.1.6.1 Results of the retailers survey.- 5.1.6.2 Results from the producers and wholesalers survey.- 5.1.7 MIS managers opinion on general standardization issues.- 5.2 Closing the standardization gap.- 5.2.1 Solution designs for standardization problems.- 5.2.2 Problems of centralized coordination.- 5.2.2.1 Data problem.- 5.2.2.2 Complexity problem.- 5.2.2.3 Implementation problem.- 5.2.3 Excursion: A word on centralization and decentralization.- 5.3 The implementation problem.- 5.3.1 A case of X.500 Directory Services.- 5.3.1.1 Directory Services.- 5.3.1.2 Empirical data.- 5.3.1.3 Profitability analysis.- 5.3.2 A network ROI.- 5.3.2.1 A two player solution.- 5.3.2.2 An n-player solution.- 5.3.3 A network ROI for a virtual principal.- 5.3.4 Problems associated with the ROI.- 5.4 The data problem.- 5.4.1 An adapted Groves mechanism for the decentralized standardization model.- 5.4.1.1 The Groves mechanism.- 5.4.1.2 A Groves example.- 5.4.1.3 A Groves mechanism for decentralized standardization problems.- 5.4.1.4 Example of dominant strategies for the adapted Groves mechanism.- 5.4.1.5 Summary.- 5.4.2 Problems associated with the Groves mechanism.- 5.5 A bidding mechanism.- 5.5.1 Two examples.- 5.5.2 A comparison between the bidding mechanism and centralized coordination.- 5.6 Managerial and policy implications.- 5.6.1 Some general findings.- 5.6.2 Networks as a competitive advantage.- 5.6.3 Committees and consortia.- 5.6.4 Asymmetric network costs and gains.- 5.6.5 Infrastructure subsidies and SME integration.- 6 Theoretical implications: towards an interdisciplinary network theory.- 6.1 General drawbacks of the neo-classical paradigm.- 6.1.1 Absence of externalities.- 6.1.2 Complete rationality of the homo oeconomicus.- 6.1.3 Exclusion principle.- 6.1.4 Consumption paradigm.- 6.1.5 Separation of consumers and producers.- 6.1.6 Divisibility of resources.- 6.1.7 Concave utility functions / no complementarities.- 6.1.8 Absence of transaction costs.- 6.2 Towards an interdisciplinary theory of network effects.- 6.3 Required modeling power of an interdisciplinary theory of network effects.- 6.3.1 Modeling of knowledge and uncertainty/bounded rationality.- 6.3.2 Evolutionary system dynamics.- 6.3.3 Emergence of system components and links.- 6.3.4 Abolishment of convexity and divisibility assumptions.- 6.3.5 Economics of intermediation.- 7 Conclusions and further research.- 7.1 Summary of the findings.- 7.2 Further research.- Variables and symbols.- List of equations.- List of figures.- List of tables.- References.

Mining Networks and Central Entities in Digital Libraries. A Graph Theoretic Approach Applied to Co-author Networks

Abstract: Based on graph theoretic concepts, the paper introduces the notion of a social network for mining central entities in bibliographic Digital Libraries. Due to this issue, several concepts of centrality in social networks have been employed that evaluate the strategic position of actors in a social network. Moreover, the paper discusses a structural technique (MPA*) that optimizes network propagation and the evaluation of actor centrality without imposing low depth threshold. The models proposed have been applied to co–author networks for finding human experts in scientific domains. The expressiveness of our network model is demonstrated by testing the precision of document sets that have been ranked by author centrality.

Network Effects and Diffusion Theory: Network Analysis in Economics

Abstract: In this paper, some of the main results of an interdisciplinary research project on standards are presented and integrated into a single framework of technology diffusion. Based on network effect theory and diffusion theory, we present an agent-based simulation model that extends the traditional economical network perspective by incorporating structural determinants of networks (centrality, topology/density) from sociology and geography and individual decision making on the part of prospective technology users under realistic informational assumptions. Based upon these models, decision behavior in terms of the selection of standards and the diffusion of technological innovations in networks can be described. The model has so far served as both a tool for developing and evaluating network strategies in practical application like EDI networks or corporate directory service planning as well as for providing theoretical insights into standardization problems and possible solutions.

Wanted: A Good Network Theory of Policy Making 1

Abstract: In the policy literature, networks have now for some time been named as an important governance mechanism. From the level of global governance, European integration, and sectoral policy networks to regional arrangements, policy networks are seen as a way to integrate differentiated actor systems and to adjust to problems that cannot be tackled by existing formal institutional configurations. Numerous empirical cases of policy networks have been produced and the development and structures of policy networks have been explained, but what seems missing is theory building on the role of networks in policy making. Therefore, in the paper we will first introduce Wallace’s model for theory development. Secondly, we will apply this model and evaluate the literature according to the question what has been achieved so far. We will then suggest what steps should be taken next in order to make some progress towards the development of a network theory of policy making.

Network Analysis, Social Dynamics and Feedback in Social Systems

Abstract: Network analysis, an area of mathematical anthropology and sociology crucial to the linking of theory and observation, developed dramatically in recent decades. This made possible a new understanding of social dynamics as a synthesis of network theories. Concrete links can be identified between the actions of self-reflective agents, with rich information processing and decision processes deeply embedded in social worlds, and emergence or change in the self-restructuring systems they operate – including the emergence of organizations, groups, institutions, norms and cultures. The last four decades saw a massive development of concepts and tools for network analysis, initially spurred by anthropologist Clyde Mitchell and sociologist Harrison White. These developments enabled burgeoning applications to ever-wider sets of problems in the social sciences. The trajectories of social network analysis in the two disciplines, however, were very different. Network approaches in sociology became rather quickly a central theoretical paradigm for integrating the dynamics of human agency with theories of the feedback between structural constraints and the emergence of institutions out of human interaction (Mullins 1973, Berkowitz 1982, Burt 1982, Wellman 2000). Still, even in sociology, the development of methodology (Wasserman and Faust 1994) far outstripped that of an integrated theory of networks that situates explanatory principles in a common conceptual framework. The anthropology of the 1960s failed to see the more general relevance of an array of network modeling possibilities to social theory, and relegated network analysis to a ‘toolkit’ for specialized problems collateral to but mostly outside of institutional and cultural analysis. 1 Interest in networks largely died out in anthropology by the mid-1970s once those who had begin experimenting with the approach in the 1960s turned from problems of fluid social structure to the study of transactions, ritual enactment, symbolic action, and contemporary themes of cultural anthropology. Anthropologists with a cognitive focus narrowed their studies to the shared components of egocentric cognitive constructions in relation to observed behavior, studies that unfortunately didn’t recapture the interests of the field at large. In the past decade, however, the effort that anthropologists have put into long-term field sites has begun to pay off in terms of longitudinal network studies of the dynamics of social networks (see Johanson and White in Kemper and Royce 2002). The rich ethnographic context that long-term field site data bring to network analysis has begun to contribute in major ways to foundational theory in the social sciences. Studies in this context have begun to integrate, in an emergent network theory, “models of how complex, information processing, self-reflective, self-restructuring systems operate, develop and change” (Read 1990: 55). Network theory has helped to formulate general explanatory frameworks to understand how multiple types of phenomena are linked and feedback on one another through their embeddings in multiple overlapping and interpenetrating network configurations. Rich ethnographic groundings for the study of multiple embedded network processes have begun to provide breakthroughs in the study of feedback processes (White and Houseman 2002). I will focus here on applications used to date for several long-term field site analyses. 1 Conceptual Perspective One of the key ingredients of scientific explanation and the testing of theory is the development of models that relate first principles, as a function of measurable parameters of interaction and structure, to a diversity of observable outcomes. Network theory generally, in so doing, attempts to explicate how social and cultural phenomena emerge out of interaction (such outcomes are path dependent). This may be done by measuring the properties of local interactions as well as different kinds of emergent structure, ideally through time, across observable networks of communication and of social and instrumental relations, events and activities. Table 1 shows some of the network concepts applicable to domains of social theory. Coupled with the modeling of fundamental interaction processes, they are designed to allow for measuring local network properties and structural emergents in order to test hypotheses about processes, interactions and outcomes. 2 The author’s PI participation in these projects were funded, most recently, under NSF grant #BCS-9978282, “Longitudinal Network Studies and Predictive Cohesion Theory,” in which Frank Harary is consultant. 2 (table 1 about here) The middle column in Table 1 lists the typical kind of mathematical model used for a particular concept. These different models can be used in combination both to build a general framework of interrelated models useful for formulating network theory, and to help test some of the hypotheses derived from network theory. 2 Structure and Dynamics

Essays on Network Effects and Money

Abstract: This thesis contains four separate papers in the field of monetary economics. The common basis for all the papers is that they deal with money and network effects in one way or another. Although the common basis is the same in all the articles, different approaches are used: theory, econometrics and history of economic thought. This first essay focuses on the network effect in the use of money and the problems associated with this. A two-currency model is constructed to derive conditions about Nash equilibria inducing monetary unions and parallel currencies. Individual differences among the population are allowed, in contrast to a representative agent model. Equilibria are derived both when rational individuals make currency choices individually and when choices are made by vote maximizing politicians. The results are compared and evaluated according to the Pareto criterion. It is shown that neither political decision, nor the market process will guarantee an efficient outcome. The political solution will, however, more often give preferable solutions. Carl Menger has been cited as a forerunner to the network theory of money. The second essay analyzes Carl Menger's monetary theory and evaluates whether he was aware of the network characteristic of money. The analysis is focused on the German word Absatzfahigkeit, which is defined and analyzed by Menger and is central in his monetary theory. An English translation of this word can be marketability or saleability. Parts of this concept cover an essential idea behind the network property of a potential medium of exchange. The result is, nevertheless, mixed. It cannot be said that his analysis is complete enough to call him a forerunner in all respects. In the third essay the changeover process from the local national currency to the Euro in the countries participating in the third stage of EMU is described and analyzed. The article tries to find factors that can help to explain the different pace of transition between the countries introducing Euro bank notes. It is found that the initial supply of Euros, the country size, the duration of dual circulation and the number of bank branches (per capita) in the economy explain a large part of the variation in the pace of adoption. The first two factors can have its origin in the network property of currency. Psychological factors, such as nationalism, have not been found to significantly influence the outcome. The final essay analyzes the adoption pattern of a new currency. The purpose of the essay is to construct two different types of models, network models and mechanical velocity models, which may help to explain the adoption path and to enrich the theories explaining the adoption process of a new currency. It is shown that network models can generate both concave, convex and S-shaped adoption functions. The velocity models can only generate concave adoption functions. Both types of models can be used to derive the observed concave adoption pattern during the currency exchange period experienced in all Euro area countries.

Examining Networks of Building Professionals, Developers, Owners and Contractors in the Commercial Building Sector

Abstract: Social networks are the key to understanding the diffusion of ideas and products in the market place. However, there have been few studies that have employed network concepts and methods to study energy efficiency programs and energy markets. In this paper, the history of network theory is briefly presented. Particular emphasis is placed on recent developments that have led to an increased interest in network science and a large amount of new work since 1998. Two examples of the analysis of networks of players in the commercial buildings market are presented. These analyses are based on F. W. Dodge Players Data for New Jersey for 2000 and 2001. The preparation of the data and the use of UCINET6 and PAJEK, network analysis tools, are discussed. One set of results identifies the architect and engineering firms associated with Limited Brands, a national retailer, and illustrates the concept of “image architects” who do much of the design and implementation of buildings for national and regional chains. The second example identifies a cluster of local firms that deal with public buildings in New Jersey. The application of network analysis has substantial potential for improving the effectiveness of program implementation efforts and the measurement of the effects of market transformation activities.

Sales channels: a barrier to entry in the airline industry

Abstract: Low cost airlines have attracted new kinds of customers and travels, created new market segments and thus expanded the total potential market This industry growth is hindered by several factors This paper examines if the distribution channel is one factor that can prohibit new entries into the market The distribution strategies chosen by new low cost airlines are studied, given two options: the travel agent and the Internet The paper also examines if the distribution of ticket sales could constitute an entrance problem Empirical data is gathered from four major airlines operating in the US and four airlines operating in Sweden The interviews were conducted in 1999 and 2003 Analysis is made using network theory, where actors in the network contribute with resources of different relative strengths and create dependencies through linkages, and channel theory Preliminary findings show that there are at least five different ways airlines can position themselves in the distribution channel network: through traditional agent, agent Web, computer reservation system Web, proprietary Web or collaborative Web Which alternative or alternatives the new entrant should choose is not self-evident Once chosen, there seems to be an entrance problem when the new airline tries to get a position, since it does not yet have a given place in the distribution network

Development of the Market Paradigm for Analyzing Systems

Abstract: The economy is heterogeneous, complex, dynamic, interdependent and interactive. This new market paradigm developed for analyzing economic behaviors gives a holistic analysis of the system by employing the Behavioral Systems Approach, Cobweb Theorem and Evolutionary Game Theory in obtaining the operating characteristics of the system through the derivation of motivation, appreciation and inertia which give the information for universal differences in the economic system. These three indices reveal, through the six marketing scenarios generated, the dynamic process that governs the evolution of the individual behavior inherent in the system. These results can explain all possible types of behavior inherent in any economic system. The computational method finds application in aggregation of heterogeneous agents, the analysis of cooperation, bounded rationality, interactive particle systems and economics, economic graphs and network theory, etc. It has also been used in analyzing conflict in an agricultural system, understanding stress-strain economics and strategic debt management in systems.

Social networks in fish

Abstract: Social network theory underwent rapid development in the 1970s by sociologists and was recently further refined by physicists. Network theory has been applied to systems consisting of interconnected components as different as that of actors in Hollywood, the neural network of the nematode worm Caenorhabditis elegans, and the US national power grid. Network theory has great potential as a conceptual tool, for example, it can be used to provide quantitative predictions regarding information transmission, spreading of diseases, and the potential for establishing reciprocal altruistic relationships. Conventionally behavioural studies have been restricted to the analysis of social interactions between isolated pairs or small groups of individuals. Here we built up, however, from pair-wise interactions, the social network of an entire population of guppies (Poecilia reticulata) in Trinidad. We show that guppies have highly structured social networks with significant repeated interactions between male–male, female–female and male–female pairs. Furthermore, sexual segregation within the network is positively correlated with the strength of the social interactions. Finally, individuals were observed to be more variable in their inter-sexual interactions than their intra-sexual ones.

A note on network structure: functional relationships between network indicators

Abstract: Two important aspects of network structure are network density and network centrality. These two aspects highlight different characteristics of a network. Empirical studies tend to find that one or the other is the better explanatory variable for certain aspects of corporate behaviour. This note makes a simple and compelling point. There is a hill-shaped relationship between network density and network centrality. A purely stochastic process of alliance formation will lead to a hill-shaped relationship. A mathematical, functional relationship can be shown between network density and the upper bound to network centrality. The combination of these outcomes sets lower and upper bounds on network centrality. This approach can offer a vantage point for an evolutionary study of network structure.