A fundamental task in data analysis is understanding the differences between several contrasting groups. These groups can represent different classes of objects, such as male or female students, or the same group over time, e.g. freshman students in 1993 through 1998. We present the problem of mining contrast sets: conjunctions of attributes and values that differ meaningfully in their distribution across groups. We provide a search algorithm for mining contrast sets with pruning rules that drastically reduce the computational complexity. Once the contrast sets are found, we post-process the results to present a subset that are surprising to the user given what we have already shown. We explicitly control the probability of Type I error (false positives) and guarantee a maximum error rate for the entire analysis by using Bonferroni corrections.

Although computer speed has steadily increased and memory is getting cheaper, the need for storage managers to deal efficiently with applications that cannot be held into main memory is vital. Dealing with large quantities of clauses implies the use of persistent knowledge and thus, indexing methods are essential to access efficiently the subset of clauses relevant to answering a query. We introduce PerKMan, a storage manager that uses G-trees, and aims at efficient manipulation of large amounts of persistent knowledge. PerKMan may be connected to Prolog systems that offer an external C language interface. As well as the fact that the storage manager allows different arguments of a predicate to share a common index dimension in a novel manner, it indexes rules and facts in the same manner. PerKMan handles compound terms efficiently and its data structures adapt their shape to large dynamic volumes of clauses, no matter what the distribution. The storage manager achieves fast clause retrieval and reasonable use of disk space.

Although Data Mining (DM) may often seem a highly effective tool for companies to be using in their business endeavors, there are a number of pitfalls and/or barriers that may impede these firms from properly budgeting for DM projects in the short term. This chapter indicates that the pitfalls of DM can be categorized into several distinct categories. We explore the issues of accessibility and usability, affordability and efficiency, scalability and adaptability, systematic patterns vs. sample-specific patterns, explanatory factors vs. random variables, segmentation vs. sampling, accuracy and cohesiveness, and standardization and verification. Finally, we present the technical challenges regarding the pitfalls of DM.

Amalthaea is an evolving, multi-agent ecosystem for personalized filtering, discovery, and monitoring of information sites. Amalthaea's primary application domain is the World Wide Web and its main purpose is to assist its users in finding interesting information. Two different categories of agents are introduced in the system: filtering agents that model and monitor the interests of the user and discovery agents that model the information sources.A market-like ecosystem where the agents evolve, compete, and collaborate is presented: agents that are useful to the user or other agents reproduce, while low-performing agents are destroyed. Results from various experiments with different system configurations and varying ratios of user interests versus agents in the system are presented. Finally issues like fine-tuning the initial parameters of the system and establishing and maintaining equilibria in the ecosystem are discussed.

An information system is a socio-technical discipline involving both human and non-human entities. Much of the research done in an information system context investigates changes caused by the introduction of new business or organisational system, or changes made to an existing system, and so can be regarded as research into aspects of technological innovation. Information systems are complex entities and their development is a complex undertaking. Research in information systems needs to take account of the complexity of information systems development rather than take steps to hide this. An approach to information systems research, based on actor-network theory, offers a good means of allowing impartial treatment of the contributions of both human and non-human actors, and of handling the complexities involved. This chapter outlines such an approach.

Analysis and design of Information Systems (ISs) is the process of eliciting the system's requirements and transforming them into a model that can be used to develop ISs. Analysis and design of Agent-Oriented Information Systems (AOISs) relates to the very same process using the multi-agent paradigm. A comprehensive and rigorous methodology for developing multi-agent systems is lacking (Elammari & Lalonde, 1999; Odell et al., 2000). Most existing multi-agent systems were developed in an ad-hoc manner, and systems developers paid little attention to requirements specification and the analysis process (Treur, 1999a).The paper has two goals: (a) to provide an overview and (b) to discuss challenges and future research of the field. To address the first goal, we review different methodologies that are suitable for analysing and designing AOIS. This is done by examining, for each methodology, its suitability in supporting the early phases of the software engineering process (specifically analysis and design) as well as its capabilities for modelling agent-oriented systems. To address the second goal, we analyse the limitations of existing approaches, identify critical issues and point to what we think are possible future directions.

As they are maturing�i.e., as they are becoming validated, calibrated and refined�computational emulation models of organizations are evolving into: powerful new kinds of organizational design tools for predicting and mitigating organizational risks; and flexible new kinds of organizational theorem-provers for validating extant organization theory and developing new theory. Over the past 50 years, computational modeling and simulation have had enormous impacts on the rate of advancement of knowledge in fields like physics, chemistry and, more recently, biology; and their subsequent application has enabled whole new areas of engineering practice. In the same way, as our young discipline comes of age, computational organizational models are beginning to impact behavioral, organizational and economic science, and management consulting practice. This paper attempts to draw parallels between computational modeling in natural sciences and computational modeling of organizations as a contributor to both social science and management practice. To illustrate the lifecycle of a computational organizational model that is now relatively mature, this paper traces the evolution of the Virtual Design Team (VDT) computational modeling and simulation research project at Stanford University from its origins in 1988 to the present. It lays out the steps in the process of validating VDT as a �computational emulation� model of organizations to the point that VDT began to influence management practice and, subsequently, to advance organizational science. We discuss alternate research trajectories that can be taken by computational and mathematical modelers who prefer the typical natural science validation trajectory�i.e., who attempt to impact organizational science first and, perhaps subsequently, to impact management practice. The paper concludes with a discussion of the current state-of-the-art of computational modeling of organizations and some thoughts about where, and how rapidly, the field is headed.

Author cocitation analysis was used to explore ongoing changes in the intellectual structure of the hybrid problem area of developmental dyslexia for the period 1994-1998, and to address ambiguities in results raised by an earlier study of these researchers for the years 1976-1993. Results suggest that: (1) discrepancies between the structure of the sociometric (personal) and author cocitation networks reflect real differences, not temporal factors; (2) differences between cocitation patterns and reports in the literature, and corresponding delays in the visibility of emerging perspectives, are likely due to the "inertia" of aggregate cocitation data and/or by shifts by neuroscience-vision researchers to publication in more prominent journals; (3) a sharp rise in link density for the neuroscience-vision subgroup indicates increased cohesiveness and growing maturation for this emerging perspective; (4) shifts in subgroup membership, link density, patterns of coauthorship, and multiple factor loadings suggest possible convergence between other subgroups in the network and identify individuals who may play boundary-spanning roles within the network; and (5) changing patterns of cocitation throughout the network suggest the increasing influence of studies relating to neurobiological mechanisms underlying dyslexia. The possible contributions of such boundary spanners in addressing the substantial information and communication challenges posed by the increased interdisciplinary character of scholarship in general also are discussed.

Awareness of others' interests can lead to fruitful collaborations, friendships and positive social change. Interviews of groups involved in both research and corporate work revealed a lack of awareness of shared interests among workers sharing an organizational affiliation and collocated in the same building or complex but still physically separated (e.g., by walls or floors). Our study showed that loosely coupled groups were less likely to discover shared interests in the way that many tightly collocated groups do, such as by overhearing conversations or noticing paraphernalia. Based on these findings we iteratively developed a system to capture and display shared interests. Our platform includes an e-mail sensor to discover personal interests, a search algorithm to determine shared interests, a public peripheral display and lightweight location-tracking system to convey those interests. We deployed the system to two groups for two months and found that the system did lead to increased awareness of shared interests.

Based on Durkheim's idea that social differentiation in modern societies leads to division of labour and to increasing alienation between individuals some authors argue that the use of new technologies like the Internet will promote social isolation. Such a tendency towards a decline in Social Capital has been reported for the U.S. while citing decreasing numbers of membership in diverse organisations. The paper investigates similar tendencies for Switzerland and discusses the appropriateness of respective measures of social capital.

Communicating scientists' behavior, as well as their ideas, computer-supported cooperative work technology fosters virtual social networks of far-flung collaborators pursuing mutual interests and experiments.

Communication technologies support virtual R&D groups by enabling immediate and frequent interaction of their geographically-distributed members. Performance of members in such groups has yet to be studied longitudinally. A model proposes not only direct effects of functional role, status, and communication role on individual performance, but also indirect effects through individual centrality. Social network analysis was performed on e-mail samples from two time periods separated by four years. Analysis revealed both direct and indirect effects as hypothesized; however, the indirect effects were more consistent in both time periods. The clearest findings were that centrality mediates the effects of functional role, status, and communication role on individual performance. Interestingly, centrality was a stronger direct predictor of performance than the individual characteristics considered in this study. The study illustrates the usefulness of accounting for network effects for better understanding individual performance in virtual groups.

Communications of AIS Volume 1, 1999 Article 7 2 Knowledge Management Systems: Issues, Challenges, and Benefits by Alavi and Leidner The knowledge-based theory of the firm suggests that knowledge is the organizational asset that enables sustainable competitive advantage in hypercompetitive environments. The emphasis on knowledge in today�s organizations is based on the assumption that barriers to the transfer and replication of knowledge endow it with strategic importance. Many organizations are developing information systems designed specifically to facilitate the sharing and integration of knowledge. Such systems are referred to as Knowledge Management System (KMS). Because KMS are just beginning to appear in organizations, little research and field data exists to guide the development and implementation of such systems or to guide expectations of the potential benefits of such systems. This study provides an analysis of current practices and outcomes of KMS and the nature of KMS as they are evolving in fifty organizations. The findings suggest that interest in KMS across a variety of industries is very high, the technological foundations are varied, and the major concerns revolve around achieving the correct amount and type of accurate knowledge and garnering support for contributing to the KMS. Implications for practice and suggestions for future research are drawn from the study findings.

Communities with high levels of social capital are likely to have a higher quality of life than communities with low social capital (Coleman, 1988, 1990; Putnam, 1993, 2000). This is due to the greater ability of such communities to organize and mobilize effectively for collective action because they have high levels of social trust, social networks, and well-established norms of mutuality (the major features of social capital). Communities with 'bridging' social capital (weak ties across groups) as well as 'bonding' social capital (strong ties within groups) are the most effective in organizing for collective action (Granovetter, 1973; Putnam, 2000). People who belong to multiple groups act as bridging ties Simmel [1908] 1950; Wellman, 1988). When people with bridging ties use communication media, such as the Internet, they enhance their capability to educate community members, and organize, as needed, for collective action. This paper summarizes evidence from stratified household survey data in Blacksburg, Virginia showing that people with weak (bridging) ties across groups have higher levels of community involvement, civic interest and collective efficacy than people without bridging ties to groups. Moreover, heavy Internet users with bridging ties have higher social engagement, use the Internet for social purposes, and have been attending more local meetings and events since going online than heavy Internet users with no bridging ties. These findings may suggest that the Internet - in the hands of bridging individuals -- is a tool for maintaining social relations, information exchange, and increasing face-to-face interaction, all of which help to build both bonding and bridging social capital in communities.

Consider a decentralized agent-based approach for service location, where agents provide and consume services, and also cooperate with each other by giving referrals to other agents. That is, the agents form a referral network. Based on feedback from their users, the agents judge the quality of the services provided by others. Further, based on the judgments of service quality, the agents also judge the quality of the referrals given by others. The agents can thus adaptively select their neighbors in order to improve their local performance. The choices by the agents cause communities to emerge. According to our definition, an agent belongs to a community only if it has been useful to the other members of the community in prior interactions regarding a particular topic. Hence, the membership in different communities is determined based on relationships among the agents. This paper compares topic-sensitive communities of the above kind with communities as studied in traditional link analysis. It studies the correlation between the two kinds of communities as they emerge in referral networks and evaluates the two kinds of communities in terms of their effectiveness in locating service providers.

Current research on distributed knowledge processes suggests a critical conflict between knowledge processes in groups and the technologies built to support them. The conflict centers on observations that authentic and efficient knowledge creation and sharing is deeply embedded in an interpersonal face to face context, but that technologies to support distributed knowledge processes rely on the assumption that knowledge can be made mobile outside these specific contexts. This conflict is of growing national importance as work patterns change from same site to separate site collaboration, and millions of government and industrial dollars are invested in establishing academic-industry alliances and building infrastructures to support distributed collaboration and knowledge. ?In this paper we describe our multi-method approach for studying the tension between embedded and mobile knowledge in a project funded by the National Science Foundation’s program on Knowledge and Distributed Intelligence. This project examines knowledge processes and technology in distributed, multidisciplinary scientific teams in the National Computational Science Alliance (Alliance), a prototypical next generation enterprise. First we review evidence for the tension between embedded and mobile knowledge in several research literatures. Then we present our three-factor conceptualization that considers how the interrelationships among characteristics of the knowledge shared, group context, and communications technology contribute to the tension between embedded and mobile knowledge. Based on this conceptualization we suggest that this dichotomy does not fully explain distributed multidisciplinary knowledge processes. Therefore we propose some alternate models of how knowledge is shared. We briefly introduce the setting in which we are studying distributed knowledge processes and finally, we describe the data collection methods and the current status of the project.

Data mining is concerned with finding interesting patterns in data. Many techniques have emerged for analyzing and visualizing large volumes of data. What one finds in the technical literature are mostly success stories of these techniques. Researchers rarely report on steps leading to success, failed attempts, or critical representation choices made; and rarely do papers include expert evaluations of achieved results. An interesting point of investigation is also why some particular solutions, despite good performance, were never used in practice or required additional treatment before they could be used. Insightful analyses of successful and unsuccessful applications are crucial for increasing our understanding of machine learning techniques and their limitations. The UCI Repository of Machine Learning Databases (Blake & Merz, 1998) has served the machine learning community for many years as a valuable resource. It has benefited the community by allowing researchers to compare algorithm performance on a common set of benchmark datasets, most taken from real-world domains. However, its existence has indirectly promoted a very narrow view of real-world data mining. Performance comparisons, which typically focus on classification accuracy, neglect important data mining issues such as data understanding, data preparation, selection of appropriate performance metrics, and expert evaluation of results. Furthermore, because the UCI repository has been used so extensively, some researchers have claimed that our algorithms may be �overfitting the UCI repository�. Challenge problems such as the KDD Cup, CoIL and PTE challenges have also become popular in recent years and have attracted numerous participants. Contrary to the �UCI challenge� of achieving best accuracy results in many different domains, these challenge problems usually involve a single difficult problem domain, and participants are evaluated by howwell their entries satisfy a domain expert. Such challenges can be a very useful source of feedback to the research community, provided that thorough analysis of results has been performed (for example, Elkan, 2001, describes lessons from the CoIL Challenge 2000). With this background in mind, a workshop on Data Mining Lessons Learned (DMLL- 2002) was organized at the Nineteenth International Conference on Machine Learning (ICML-2002) in Sydney in July of 2002 in order to gather and extract lessons from data mining endeavors. This workshop featured three invited talks and ten contributing authors presenting the different aspects of practical data mining applications and data mining competitions, together with their lessons learned. These reports are available in the on-line DMLL-2002 proceedings (Lavra?c, Motoda, & Fawcett, 2002). In early 2003 a call for papers was issued for a special issue of the Machine Learning journal. In contrast to a previous special issue of Machine Learning (volume 30, issue 2�3) on applications, the main goal of this special issue is to focus on the lessons learned from the data mining process rather than on the applications themselves. The result is this issue containing this editorial, one introductory paper and six contributed papers. The aim of this special issue is to gather experiences gained from data mining applications and challenge competitions, in terms of the lessons learned both from successes and from failures, from the engineering of representations for practical problems, and from expert evaluations of solutions.

Decision support systems (DSS) are becoming increasingly more critical to the daily operation of organizations. Data warehousing, an integral part of this, provides an infrastructure that enables businesses to extract, cleanse, and store vast amounts of data. The basic purpose of a data warehouse is to empower the knowledge workers with information that allows them to make decisions based on a solid foundation of fact. However, only a fraction of the needed information exists on computers; the vast majority of a firm's intellectual assets exist as knowledge in the minds of its employees. What is needed is a new generation of knowledge-enabled systems that provides the infrastructure needed to capture, cleanse, store, organize, leverage, and disseminate not only data and information but also the knowledge of the firm. The purpose of this paper is to propose, as an extension to the data warehouse model, a knowledge warehouse (KW) architecture that will not only facilitate the capturing and coding of knowledge but also enhance the retrieval and sharing of knowledge across the organization. The knowledge warehouse proposed here suggests a different direction for DSS in the next decade. This new direction is based on an expanded purpose of DSS. That is, the purpose of DSS in knowledge improvement. This expanded purpose of DSS also suggests that the effectiveness of a DSS will, in the future, be measured based on how well it promotes and enhances knowledge, how well it improves the mental model(s) and understanding of the decision maker(s) and thereby how well it improves his/her decision making.

Designing, building, and implementing an architecture for distributed knowledge discovery.

Enhancing traditional algorithmic techniques for improved pattern analysis.

Enterprise executives understand that timely, accurate knowledge can mean improved business performance. Two technologies have been central in improving the quantitative and qualitative value of the knowledge available to decision makers: business intelligence and knowledge management. Business intelligence has applied the functionality, scalability, and reliability of modern database management systems to build ever-larger data warehouses, and to utilize data mining techniques to extract business advantage from the vast amount of available enterprise data. Knowledge management technologies, while less mature than business intelligence technologies, are now capable of combining today's content management systems and the Web with vastly improved searching and text mining capabilities to derive more value from the explosion of textual information. We believe that these systems will blend over time, borrowing techniques from each other and inspiring new approaches that can analyze data and text together, seamlessly. We call this blended technology BIKM. In this paper, we describe some of the current business problems that require analysis of both text and data, and some of the technical challenges posed by these problems. We describe a particular approach based on an OLAP (on-line analytical processing) model enhanced with text analysis, and describe two tools that we have developed to explore this approach--eClassifier performs text analysis, and Sapient integrates data and text through an OLAP-style interaction model. Finally, we discuss some new research that we are pursuing to enhance this approach.

Existing theory has framed the process of information extraction and agglomeration, also referred to as the knowledge discovery (KD) process, as a series of strategic search decisions, subject to constraints, with the objective ot attaining a sufficient level of domain-specific knowledge for use in strategic planning. Supported by the experiences of firms representative of Client, Developer, and Third-party segments of the data mining (DM) community, this work provides an extension to this basic framework. The implications provided suggest a wealth of untapped opportunities in the area of KD research.

For many companies, competitiveness in e-commerce requires a successful presence on the web. Web sites are used to establish the company's image, to promote and sell goods and to provide customer support. The success of a web site affects and reflects directly the success of the company in the electronic market. In this study, we propose a methodology to improve the �success� of web sites, based on the exploitation of navigation pattern discovery. In particular, we present a theory, in which success is modelled on the basis of the navigation behaviour of the site's users. We then exploit WUM, a navigation pattern discovery miner, to study how the success of a site is reflected in the users' behaviour. With WUM we measure the success of a site's components and obtain concrete indications of how the site should be improved. We report on our first experiments with an online catalog, the success of which we have studied. Our mining analysis has shown very promising results, on the basis of which the site is currently undergoing concrete improvements.

For many purposes, the Web page is too small a unit of interaction and analysis. Web sites are structured multimedia documents consisting of many pages, and users often are interested in obtaining and evaluating entire collections of topically related sites. Once such a collection is obtained, users face the challenge of exploring, comprehending and organizing the items. We report four innovations that address these user needs: (1) we replaced the Web page with the Web site as the basic unit of interaction and analysis;(2) we defined a new informationstructure, the clan graph, that groups together sets of related sites; (3) we augment the representation of a site with a site profile, information about site structure and content that helps inform user evaluation of a site; and (4) we invented a new graph visualization, the auditorium visualization, that reveals important structural and content properties of sites within a clan graph. Detailed analysis and user studies document the utility of this approach. The clan graph construction algorithm tends to filter out irrelevant sites and discover additional relevant items. The auditorium visualization, augmented with drill-down capabilities to explore site profile data, helps users to find high-quality sites as well as sites that serve a particular function.

In Knowledge Discovery in Databases (KDD)/Data Mining literature, �interestingness� measures are used to rank rules according to the �interest� a particular rule is expected to evoke. In this paper, we introduce an aspect of subjective interestingness called �item-relatedness�. Relatedness is a consequence of relationships that exist between items in a domain. Association rules containing unrelated or weakly related items are interesting since the co-occurrence of such items is unexpected. �Item-Relatedness� helps in ranking association rules on the basis of one kind of subjective unexpectedness. We identify three types of item-relatedness � captured in the structure of a �fuzzy taxonomy� (an extension of the classical concept hierarchy tree). An �item-relatedness� measure for describing relatedness between two items is developed by combining these three types. Efficacy of this measure is illustrated with the help of a sample taxonomy. We discuss three mechanisms for extending this measure from a two-item set to an association rule consisting of a set of more than two items. These mechanisms utilize the relatedness of item-pairs and other aspects of an association rule, namely its structure, distribution of items and item-pairs. We compare our approach with another method from recent literature.

In the last few years, there has been increasing interest from the agent community in the use of techniques from decision theory and game theory. Our aims in this article are firstly to briefly summarize the key concepts of decision theory and game theory, secondly to discuss how these tools are being applied in agent systems research, and finally to introduce this special issue of Autonomous Agents and Multi-Agent Systems by reviewing the papers that appear.

In the last years, the main orientation of formal concept analysis (FCA) has turned from mathematics towards computer science. This article provides a review of this new orientation and analyses why and how FCA and computer science attracted each other. It discusses FCA as a knowledge representation formalism using five knowledge representation principles provided by Davis et al. (1993). It then studies how and why mathematics-based researchers got attracted by computer science. We will argue for continuing this trend by integrating the two research areas FCA and ontology engineering.The second part of the article discusses three lines of research which witness the new orientation of FCA: FCA as a conceptual clustering technique and its application for supporting the merging of ontologies; the efficient computation of association rules and the structuring of the results; and the visualization and management of conceptual hierarchies and ontologies including its application in an email management system.

In their 1998 article "Visualizing a discipline: An author cocitation analysis of information science, 1972-1995," White and McCain used multidimensional scaling, hierarchical clustering, and factor analysis to display the specialty groupings of 120 highly-cited ("paradigmatic") information scientists. These statistical techniques are traditional in author cocitation analysis (ACA). It is shown here that a newer technique, Pathfinder Networks (PFNETs), has considerable advantages for ACA. In PFNETs, nodes represent authors, and explicit links represent weighted paths between nodes, the weights in this case being cocitation counts. The links can be drawn to exclude all but the single highest counts for author pairs, which reduces a network of authors to only the most salient relationships. When these are mapped, dominant authors can be defined as those with relatively many links to other authors (i.e., high degree centrality). Links between authors and dominant authors define specialties, and links between dominant authors connect specialties into a discipline. Maps are made with one rather than several computer routines and in one rather than many computer passes. Also, PFNETs can, and should, be generated from matrices of raw counts rather than Pearson correlations, which removes a computational step associated with traditional ACA. White and McCain's raw data from 1998 are remapped as a PFNET. It is shown that the specialty groupings correspond closely to those seen in the factor analysis of the 1998 article. Because PFNETs are fast to compute, they are used in AuthorLink, a new Web-based system that creates live interfaces for cocited author retrieval on the fly.

In this paper we examine an emerging class of systems that link People-to-People-to-Geographical-Places; we call these P3-Systems. Through analyzing the literature, we have identified four major P3-System design techniques: People-Centered systems that use either absolute user location (e.g. Active Badge) or user proximity (e.g. Hocman) and Place-Centered systems based on either a representation of people's use of physical spaces (e.g. ActiveMap) or on a matching virtual space that enables online interaction linked to physical location (e.g. Geonotes). In addition, each feature can be instantiated synchronously or asynchronously. The P3-System framework organizes existing systems into meaningful categories and structures the design space for an interesting new class of potentially context-aware systems. Our discussion of the framework suggests new ways of understanding and addressing the privacy concerns associated with location aware community system and outlines additional socio-technical challenges and opportunities.

In this paper we outline a new method for clustering that is based on a binary representation of data records. The binary database relates each entity to all possible attribute values (domain) that entity may assume. The resulting binary matrix allows for similarity and clustering calculation by using the positive (�1� bits) of the entity vector. We formulate two indexes: Pair Similarity Index (PSI) to measure similarity between two entities and Group Similarity Index (GSI) to measure similarity within a group of entities. A threshold factor for each attribute domain is defined that is dependent on the domain but independent of the number of entities in the group. The similarity measure provides simplicity of storage and efficiency of calculation. A comparison of our similarity index to other indexes is made. Experiments with sample data indicate a 48% improvement of group similarity over standard methods pointing to the potential and merit of the binary approach to clustering and data mining.

In this paper, we develop a dynamic supernetwork framework for the modelling and analysis of supply chains with electronic commerce that also includes the role that relationships play. Manufacturers are assumed to produce homogeneous products and to sell them either through physical or electronic links to retailers and/or directly to consumers through electronic links. Retailers, in turn, can sell the products through physical links to consumers. Increasing relationship levels in our framework are assumed to reduce transaction costs as well as risk and to have some additional value for both sellers and buyers. Establishing those relationship levels incurs some costs that have to be borne by the decisionmakers in the supernetwork, which is multilevel in structure and consists of the supply chain and the social network. The decision-makers, who are located at distinct tiers in the supernetwork, try to optimize their objective functions and are faced with multiple criteria including relationship-related ones and weight them according to their preferences. We establish the optimality conditions for the manufacturers, retailers, and consumers, derive the equilibrium conditions, and provide the variational inequality formulation. We then present the projected dynamical system, which describes the disequilibrium dynamics of the product transactions, relationship levels, and prices on the supernetwork, and whose set of stationary points coincides with the set of solutions of the variational inequality problem. We also illustrate the dynamic supernetwork model through several numerical examples, for which the explicit equilibrium patterns are computed.

Information landscaping--an integration of information mapping, charting, querying and reporting techniques--has been developed to enable the construction of a total quality knowledge management system focusing on a particular subject information field. The techniques apply five major parameters of the Fuzzy commonality model (FCM) including unionization, quantity, continuity or stability, changeability, and critical probability, to construct a series of information maps (infomaps) and a set of chronological-statistical charts (infocharts). The infomaps and infocharts are used as the blueprints and navigation agents for building and developing a web-based subject experts depository and query-report system. Focusing on the subject experts/expertise, this system enables a researcher to expedite a query search through infomaps (qualitative reference) and infocharts (quantitative reference). The entropy measurement and the entropy constant (the square root of the average entropy measure) are calculated to compare with the critical probability of the FCM. This leads to the finding of a set of regression straight lines and the establishment of an information oscillogram. The tropics (upper limit, middle range, lower limit), and the potential/ solstitial population and its growth rate within a subject information domain during a particular time period can be determined. They can effectively and efficiently guide librarians and information professionals towards the construction and the continuous development of an electronic collection. The cultivation of a virtual learning and referencing environment can also be created by utilizing this data.

Interaction protocols are widely recognized as an essential mechanism for coordination within multi-agent systems. There is thus a need for coordination models for specifying, validating, and implementing protocols, possibly open and concurrent, efficiently and reliably. This paper proposes such a model, which considers protocols as resources and each conversation among agents following the rules of a protocol as a well-identified process. To this end, a new kind of middle-agent, called Moderator, is introduced. A Moderator is in charge of monitoring a conversation so that it progresses according to the protocol rules, and provides agents with services to ease their involvement in the conversation. This model fits the organization-centered view of multi-agent systems as it strictly distinguishes the agent-level and the organization-level concerns with regard to interaction. In addition, the paper shows that this model is supported by a High-Level Petri Net language that covers all the steps of protocol engineering: design, validation, implementation. This paper presents this Moderator Coordination Model along four related dimensions: a conceptual model of protocols, a MAS architecture, a suitable modeling formalism, and an associated development process.

Intra-organizational network research had its first heyday during the empirical revolution in social sciences before World War II when it discovered the informal group within the formal organization. These studies comment on the classic sociological idea of bureaucracy being the optimal organization. Later relational interest within organizational studies gave way to comparative studies on the quantifiable formal features of organizations. There has been a resurgence in intra-organizational networks studies recently as the conviction grows that they are critical to organizational and individual performance. Along with methodological improvements, the theoretical emphasis has shifted from networks as a constraining force to a conceptualization that sees them as providing opportunities and finally, as social capital. Because of this shift it has become necessary not only to explain the differences between networks but also their outcomes, that is, their performance. It also implies that internal and external networks should no longer be treated separately. Research on differences between intra-organizational networks centers on the influence of the formal organization, organizational demography, technology and environment. Studies on outcomes deal with diffusion and adaptation of innovation; the utilization of human capital; recruitment, absenteeism and turnover; work stress and job satisfaction; equity; power; information efficiency; collective decision making; mobilization for and outcomes of conflicts; social control; profit and survival of firms and individual performance. Of all the difficulties that are associated with intra-organizational network research, problems of access to organizations and incomparability of research findings seem to be the most serious. Nevertheless, future research should concentrate on mechanisms that make networks productive, while taking into account the difficulties of measuring performance within organizations, such as the performance paradox and the halo-effect.

Knowledge and knowledge management are receiving tremendous interest from both practitioners and academics. Although knowledge management is often accepted as a very useful organizational activity, a number of the assumptions underlying knowledge management have not been investigated. This paper examines four knowledge management assumptions: knowledge is worth managing, organizations benefit from managing knowledge, knowledge can be managed, and little risk is associated with managing knowledge. The assumptions are analyzed at strategic and operational levels, and both negating and supporting evidence is presented. Based on this analysis, a framework for research in knowledge management is proposed. The framework is used to generate a number of key questions that should be addressed in knowledge management research. Particular attention is given to goals and rewards as well as to the role of information technology in knowledge management.

Knowledge discovery in databases is a rapidly growing field, whose development is driven by strong research interests as well as urgent practical, social, and economical needs. While the last few years knowledge discovery tools have been used mainly in research environments, sophisticated software products are now rapidly emerging. In this paper, we provide an overview of common knowledge discovery tasks and approaches to solve these tasks. We propose a feature classification scheme that can be used to study knowledge and data mining software. This scheme is based on the software's general characteristics, database connectivity, and data mining characteristics. We then apply our feature classification scheme to investigate 43 software products, which are either research prototypes or commercially available. Finally, we specify features that we consider important for knowledge discovery software to possess in order to accommodate its users effectively, as well as issues that are either not addressed or insufficiently solved yet.

Many authors have posited a social component in citation, the consensus being that the citers and citees often have interpersonal as well as intellectual ties. Evidence for this belief has been rather meager, however, in part because social networks researchers have lacked bibliometric data (e.g., pairwise citation counts from online databases), and citation analysts have lacked sociometric data (e.g., pairwise measures of acquaintanceship). In 1997 Nazer extensively measured personal relationships and communication behaviors in what we call "Globenet," an international group of 16 researchers from seven disciplines that was established in 1993 to study human development. Since Globenet's membership is known, it was possible during 2002 to obtain citation records for all members in databases of the Institute for Scientific Information. This permitted examination of how members cited each other (intercited) in journal articles over the past three decades and in a 1999 book to which they all contributed. It was also possible to explore links between the intercitation data and the social and communication data. Using network-analytic techniques, we look at the growth of intercitation over time, the extent to which it follows disciplinary or inter-disciplinary lines, whether it covaries with degrees of acquaintanceship, whether it reflects Globenet's organizational structure, whether it is associated with particular in-group communication patterns, and whether it is related to the cocitation of Globenet members. Results show cocitation to be a powerful predictor of intercitation in the journal articles, while being an editor or co-author is an important predictor in the book. Intellectualties based on shared content did better as predictors than content-neutral socialties like friendship. However, interciters in Globenet communicated more than did noninterciters.

Many researchers have investigated the social and cognitive processes underlying organizational behavior, with particular interest in understanding the interaction between the social and cognitive dimensions. Because of the widespread use of graphs as models of social and cognitive structures, these studies frequently encounter the problem of analyzing collections of graphical structures. Such analyses have used a variety of approaches to address specific aspects of such structures. However, no single unified approach has emerged that supports the several different types of analyses required. The purpose of this paper is to define such an approach, based on a mathematical model for capturing the relationships among multiple graphs, and to demonstrate its application to the investigation of social and cognitive structures in organizations.

Millions of people meet online to chat, to find like-minded people, to debate topical issues, to play games, to give or ask for information, to find support, to shop, or just to hang-out with others. They go to chat-rooms, bulletin boards, join discussion groups or they create their group using instant messaging software. Short messaging (also known as �texting�) is also gaining popularity in some parts of the world. These online social gatherings are known by a variety of names including �online community�, a name coined by early pioneers like Howard Rheingold, who describes these online communities as �cultural aggregations that emerge when enough people bump into each other often enough in cyberspace� (Rheingold, 1994, p. 57).

Mining frequent patterns in transaction databases, time-series databases, and many other kinds of databases has been studied popularly in data mining research. Most of the previous studies adopt an Apriori-like candidate set generation-and-test approach. However, candidate set generation is still costly, especially when there exist a large number of patterns and/or long patterns. In this study, we propose a novel frequent-pattern tree (FP-tree) structure, which is an extended prefix-tree structure for storing compressed, crucial information about frequent patterns, and develop an efficient FP-tree-based mining method, FP-growth, for mining the complete set of frequent patterns by pattern fragment growth. Efficiency of mining is achieved with three techniques: (1) a large database is compressed into a condensed, smaller data structure, FP-tree which avoids costly, repeated database scans, (2) our FP-tree-based mining adopts a pattern-fragment growth method to avoid the costly generation of a large number of candidate sets, and (3) a partitioning-based, divide-and-conquer method is used to decompose the mining task into a set of smaller tasks for mining confined patterns in conditional databases, which dramatically reduces the search space. Our performance study shows that the FP-growth method is efficient and scalable for mining both long and short frequent patterns, and is about an order of magnitude faster than the Apriori algorithm and also faster than some recently reported new frequent-pattern mining methods.

One of the effects of the expansion of the World Wide Web is the production of a huge amount of data, differentiated for type, available to a large number of different users. Furthermore, the constant progress of computer hardware technology in the past three decades has led to the availability of powerful computers, data collection equipments, and storage media; this technology provides a great boost to the database and information industry by allowing transaction management, information retrieval, and data analysis over massive amounts of heterogeneous data. Moreover, the explosion of Internet increases the availability of data in different formats: structured (e.g. relational), semistructured (e.g. HTML, XML) and unstructured (e.g. plain text, audio/video) data [2]. Thus, new data management systems, able to take advantage of these heterogeneous data, are emerging and will play a vital role in the information industry. Thus, heterogeneous database systems emerge and play a vital role in the information industry. Knowledge Management is concerned with the technological, economic and organizational aspects related to (i) the creation, distribution, diversification and sharing of knowledge in complex organizations and to (ii) the management of informative flows, processes and interactions with external Knowledge [8]. Figure 1 summarizes the steps (each represented on a different level of the pyramid) through which knowledge is typically extracted from basic data. The first three levels regard the management of explicit knowledge (i.e. codified, structured or semistructured and completely available). In particular, starting from the bottom, the first level is concerned with storing and exchanging "factual" knowledge, essentially corresponding to basic data. Technologies used here comprise Databases [17], Data Repositories, Archive Sharing tools and the emerging Extensible Markup Language (XML) [18]. The second level regards "conceptual knowledge" modeling, i.e. the definition of concepts and relationships among them. Such knowledge is typically represented by means of diagram-based formalisms for both information and related processes [9]. The Unified Modeling Language (UML) is currently one of the most promising modeling languages, oriented towards the specification,implementation and documentation of complex software systems, but also used for modeling company processes not strictly related to the software. The third level is concerned with organization and integration of information represented according to heterogeneous formalisms. Techniques used here are essentially those concerning Data Warehousing (DW) [10]. Data warehouses are integrated repositories of data extracted from multiple heterogeneous sources, organized under a unified schema and at a single site, in order to facilitate management and decision making. Data Warehousing technologies include data cleaning, data integration, and Online Analytical Processing (OLAP), i.e. analysis techniques based on aggregation and summarization. The highest level regards Knowledge Discovery, i.e. the uncovering of new, implicit and potentially useful knowledge from large amounts of data. The core phase of knowledge discovery is Data Mining [10], an interactive, iterative, multi-step process, comprising in particular pattern searching and eventual refinements on the basis of domain experts' knowledge. In the context of explicit knowledge management, the Extensible Markup Language takes naturally place. XML is a language for semistructured data [1, 5] of the World Wide Web Consortium (W3C) [13] which is designed to allow marking, transferring and reusing information by means of a standard method of definition of the documents structure and format. Its metalanguage features have been used in knowledge management typically for (i) the semi-automatic production of documents, (ii) the reuse of semistructured information and its integration in heterogeneous systems, (iii) the creation of knowledge maps for the organization and sharing of information. The increasing quantity of available semistructured data and the use of XML for their description and exchange discovers new reaserch themes related to management and knowledge extraction over XML data. In this scenario, our proposal consists of a system for the syntesization of XML documents that attempts to extract their semantics and to derive synthetic versions of them by means of a multidimensional interpretation [10]. In the contest of Knowledge Management, data synthesization can be regarded as a new way for knowledge extraction, by discovering and aggregating (useful) core information and by neglecting (useless) details.

Organization theories differ considerably in what they represent as the most important properties of the organizational phenomenon (Pfeffer, 1997; Scott, 1998). Sometimes organizations are seen as resource allocation mechanisms (Simon, 1991;Williamson, 1991) that can be designed to substitute markets whenever the price system cannot guarantee desirable collective outcomes (Arrow, 1974). In other circumstances organizations are depicted as the main sources of power and power differences in modern societies (Coleman, 1974; Pfeffer, 1987). Organizations have been differently viewed as containers of routines and decision rules (March et al., 2001; Nelson and Winter, 1986), as stocks of solutions available to the problem of social change (Hannan and Freeman, 1989) and as sets of codes shaping and constraining collective identities (Carroll and Hannan, 2000). Sometimes organizations are treated as complex cognitive constructions (Weick, 1969), as sets of contractual relationships (Gibbons, 2001) or as intendedly rational solutions to incentive problems (March and Simon, 1958; Milgrom and Roberts, 1992). Organizations have been frequently viewed as complex adaptive computational systems (Carley, 2002; Simon, 1969) that are socially situated (Carley, 1995) and goal-directed (Aldrich, 1999). In some other cases it has proven useful to interpret organizations as patterns of decisions emerging from quasi-random flows of problems, solutions and decision-makers (Cohen et al., 1972; Cohen and March, 1976). Institutional theories portray organizations as rationalizing agents and as �recalcitrant� tools of economic, cognitive and cultural control (Selznick, 1948; DiMaggio and Powell, 1983). One of the key unifying themes of interest to contemporary students of organizations across a variety of substantive research areas and epistemological orientations is the understanding of how different network ties concatenate to shape the evolution of distinct types of social forms and social settings. Examples of such settings include firms, markets, industries and states (Breiger, 2002; Cederman, 1997; DiMaggio, 2001; Powell et al., 1996; Rauch and Casella, 2001; White, 2002). As the papers contained in this special issue collectively demonstrate, the interest in networks has sharpened the focus on the development of increasingly sophisticated theoretical accounts of how different types of relations are implicated in a wide range of organizational processes. Examples of such processes include the emergence of new organizational forms (Padgett and Ansell, 1993; Stark, 2001), and the maintenance and erosion of individual and collective identities across levels of analysis (Breiger, 2000; Mische and Pattison, 2000). Stimulated by these broad concerns, recent years have witnessed a marked increase both in empirical network studies within and between organizations, and in theoretical speculations as to the possible consequences of networks for processes of boundary formation and dissolution around individuals, institutions and social forms. Such studies often are based on�and give rise to�complex relational data structures that call for a parallel growth in the sophistication of mathematical representation and models for the analysis of social networks and network-related processes. The aim of this special issue is to bring to the attention of members of the computational analysis of social and organizational systems community a selected number of innovative and high-quality contributions that illustrate clearly the relevance of network-based models to the study of complex organizations.

Participants in social network sites create self-descriptive profiles that include their links to other members, creating a visible network of connections � the ostensible purpose of these sites is to use this network to make friends, dates, and business connections. In this paper we explore the social implications of the public display of one's social network. Why do people display their social connections in everyday life, and why do they do so in these networking sites? What do people learn about another's identity through the signal of network display? How does this display facilitate connections, and how does it change the costs and benefits of making and brokering such connections compared to traditional means? The paper includes several design recommendations for future networking sites.

Prior studies have questioned the degree of anonymity of the double-blind review process for scholarly research articles. For example, one study based on a survey of reviewers concluded that authors often could be identified by reviewers using a combination of the author's reference list and the referee's personal background knowledge. For the KDD Cup 2003 competition's "Open Task," we examined how well various automatic matching techniques could identify authors within the competition's very large archive of research papers. This paper describes the issues surrounding author identification, how these issues motivated our study, and the results we obtained. The best method, based on discriminative self-citations, identified authors correctly 40--45% of the time. One main motivation for double-blind review is to eliminate bias in favor of well-known authors. However, identification accuracy for authors with substantial publication history is even better (60% accuracy for the top-10% most prolific authors, 85% for authors with 100 or more prior papers).

Reading recent knowledge management (KM) articles, one cannot escape the impression of a recycled concept. Definitions of the new field look remarkably like those of information systems, decision support systems, and even data management of the past. Since we believe KM is essentially new, a refined articulation of KM is desirable. Our point of departure is the observation that yesterday�s data are today�s information, which will become tomorrow�s knowledge, and knowledge, in turn, recycles down the value chain back into information and into data. We outline a framework of KM that articulates the basic terms of this perpetual process. The proposed model defines operations and transformations of data-to-information, information-to-knowledge, and their reverse order. Such transformations correspond to a time dimension of pastpresent- future and resemble the process of abstraction. Based on our analysis, we conclude that knowledge management is truly a new idea, not a recycled concept.

Recent research on small-world theory has been expanded upon organizations behavior. However, most individual studies have taken a simplified view of individuals and relationships among members by focusing on a single type of links mainly within dyadic relationships. In reality, members of organizations are interacting with each other and often connected via many types of links within more complex relationships. To explore this complex interaction, this study models organization as network and proposes a multiplex approach that captures the complexity of relationships among members in organizations. This approach accounts for the multiple types of links among members and the multiple roles of members within network organizations. It is illustrated via a case study of a network organization. The case study demonstrates how this approach could capture the many types of relationship among members as well as the various roles that members play within the network organization. Such an approach can yield new insights on how to better manage network organizations.

Recently the awareness of the importance of replicating studies has been growing in the empirical software engineering community. The results of any one study cannot simply be extrapolated to all environments because there are many uncontrollable sources of variation between different environments. In our work, we have reasoned that the availability of laboratory packages for experiments can encourage better replications and complementary studies. However, even with effectively specified laboratory packages, transfer of experimental know-how can still be difficult. In this paper, we discuss the collaboration structures we have been using in the Readers� Project, a bilateral project supported by the Brazilian and American national science agencies that is investigating replications and transfer of experimental know-how issues. In particular, we discuss how these structures map to the Nonaka�Takeuchi knowledge sharing model, a well-known paradigm used in the knowledge management literature. We describe an instantiation of the Nonaka�Takeuchi Model for software engineering experimentation, establishing a framework for discussing knowledge sharing issues related to experimental software engineering. We use two replications to illustrate some of the knowledge sharing issues we have faced and discuss the mechanisms we are using to tackle those issues in Readers� Project.

Refuting the notion of technology as a replacement of knowledge, this paper focuses on a gap between them that needs to be bridged. The idea is that technology represents the means, and knowledge the end of a process that includes many explicit and implicit methods for generating knowledge by using technology. Among these methods is data mining (DM), the leading thrust in the effort to gain actionable information from operational databases of organizations; this is particularly evident in direct marketing, customer relationship management (CRM), user profiling, and e-commerce applications.Two models of knowledge are reviewed. The first follows a conventional hierarchy of data, information and knowledge with a spiral and recursive way of generating knowledge. The other presents a reverse hierarchy where knowledge precedes the data-to-information process. The models are compared and discussed in the context of knowledge management (KM), using DM as an example.

Research in organizational learning has demonstrated processes and occasionally performance implications of acquisition of declarative (know-what) and procedural (know-how) knowledge. However, considerably less attention has been paid to learned characteristics of relationships that affect the decision to seek information from other people. Based on a review of the social network, information processing, and organizational learning literatures, along with the results of a previous qualitative study, we propose a formal model of information seeking in which the probability of seeking information from another person is a function of (1) knowing what that person knows; (2) valuing what that person knows; (3) being able to gain timely access to that person's thinking; and (4) perceiving that seeking information from that person would not be too costly. We also hypothesize that the knowing, access, and cost variables mediate the relationship between physical proximity and information seeking. The model is tested using two separate research sites to provide replication. The results indicate strong support for the model and the mediation hypothesis (with the exception of the cost variable). Implications are drawn for the study of both transactive memory and organizational learning, as well as for management practice.

Since the events of September 11, 2001, the United States has found itself engaged in an unconventional and asymmetric form of warfare against elusive terrorist organizations. Defense and investigative organizations require innovative solutions that will assist them in determining the membership and structure of these organizations. Data on covert organizations are often in the form of disparate and incomplete inferences of memberships and connections between members. NETEST is a tool that combines multi-agent technology with hierarchical Bayesian inference models and biased net models to produce accurate posterior representations of a network. Bayesian inference models produce representations of a network's structure and informant accuracy by combining prior network and accuracy data with informant perceptions of a network. Biased net theory examines and captures the biases that may exist in a specific network or set of networks. Using NETEST, an investigator has the power to estimate a network's size, determine its membership and structure, determine areas of the network where data is missing, perform cost/benefit analysis of additional information, assess group level capabilities embedded in the network, and pose �what if� scenarios to destabilize a network and predict its evolution over time.

Social action is situated in fields that are simultaneously composed of interpersonal ties and relations among organizations, which are both usefully characterized as social networks. We introduce a novel approach to distinguishing different network macro-structures in terms of cohesive subsets and their overlaps. We develop a vocabulary that relates different forms of network cohesion to field properties as opposed to organizational constraints on ties and structures. We illustrate differences in probabilistic attachment processes in network evolution that link on the one hand to organizational constraints versus field properties and to cohesive network topologies on the other. This allows us to identify a set of important new micro-macro linkages between local behavior in networks and global network properties. The analytic strategy thus puts in place a methodology for Predictive Social Cohesion theory to be developed and tested in the context of informal and formal organizations and organizational fields. We also show how organizations and fields combine at different scales of cohesive depth and cohesive breadth. Operational measures and results are illustrated for three organizational examples, and analysis of these cases suggests that different structures of cohesive subsets and overlaps may be predictive in organizational contexts and similarly for the larger fields in which they are embedded. Useful predictions may also be based on feedback from level of cohesion in the larger field back to organizations, conditioned on the level of multiconnectivity to the field.

Social data mining systems enable people to share opinions and benefit from each other's experience. They do this by mining and redistributing information from computational records of social activity such as Usenet messages, system usage history, citations, or hyperlinks. Some general questions for evaluating such systems are: (1) is the extracted information valuable? and (2) do interfaces based on the information improve user task performance? We report here on TopicShop, a system that mines information from the structure and content of Web pages and provides an exploratory information workspace interface. We carried out experiments that yielded positive answers to both evaluation questions. First, a number of automatically computable features about Web sites do a good job of predicting expert quality judgments about sites. Second, compared to popular Web search interfaces, the TopicShop interface to this information lets users select significantly more high-quality sites, in less time and with less effort, and to organize the sites they select into personally meaningful collections more quickly and easily. We conclude by discussing how our results may be applied and considering how they touch on general issues concerning quality, expertise, and consensus.

Studying and evaluating real experiences that promote active and collaborative learning is a crucial field in CSCL. Major issues that remain unsolved deal with the merging of qualitative and quantitative methods and data, especially in educational settings that involve both physical and computer-supported collaboration. In this paper we present a mixed evaluation method that combines traditional sources of data with computer logs, and integrates quantitative statistics, qualitative data analysis and social network analysis in an overall interpretative approach. Several computer tools have been developed to assist in this process, integrated with generic software for qualitative analysis. The evaluation method and tools have been incrementally applied and validated in the context of an educational and research project that has been going on during the last three years. The use of the method is illustrated in this paper by an example consisting of the evaluation of a particular category within this project. The proposed method and tools aim at giving an answer to the need of innovative techniques for the study of new forms of interaction emerging in CSCL; at increasing the efficiency of the traditionally demanding qualitative methods, so that they can be used by teachers in curriculum-based experiences; and at the definition of a set of guidelines for bridging different data sources and analysis perspectives.

Supporting mobility in a multi hop wireless environment like the MANET still remains a point of research, especially in the context of time-constrained applications. The incapacity of ad hoc networks to offer services of the likes of static or infratructured networks may be attributed to two major reasons. One, unpredictable mobility of hosts cause location-transparent-packet-delivery to be implemented only at the expense of large control overhead. Two, the lack of central control causes connection management and scalability to be major problems in the multi hop environment. In this paper we propose an efficient agent based routing mechanism that not only incurs minimal overhead, but also lays the foundation for additional functionalities as network management and real time applications. In other words, we show that the agent framework makes the MANET robust and survivable under stringent system constraints.

Teamwork during IS development (ISD) is an important issue. This paper discusses the relationship betweer team structure and ISD team performance using a social network approach. Based on empirical evidence collected from 125 teams in a system analysis and design course, we found that: (1) Group cohesion was positively related to overall performance. (2) Group conflict indexes were not significantly correlated with overall performance. (3) Group characteristics, e.g., cohesion and conflict, fluctuated in different phases, but in later stages, much less cohesion occurred and the advice network seemed to be very important. (4) Group structures seemed to be a critical factor for good performance.Further in-depth studies were conducted on teams exhibiting the highest and lowest performance to determine their differences from a sociogram analysis perspective.

The development of the semantic Web will require agents to use common domain ontologies to facilitate communication of conceptual knowledge. However, the proliferation of domain ontologies may also result in conflicts between the meanings assigned to the various terms. That is, agents with diverse ontologies may use different terms to refer to the same meaning or the same term to refer to different meanings. Agents will need a method for learning and translating similar semantic concepts between diverse ontologies. Only until recently have researchers diverged from the last decade's �common ontology� paradigm to a paradigm involving agents that can share knowledge using diverse ontologies. This paper describes how we address this agent knowledge sharing problem of how agents deal with diverse ontologies by introducing a methodology and algorithms for multi-agent knowledge sharing and learning in a peer-to-peer setting. We demonstrate how this approach will enable multi-agent systems to assist groups of people in locating, translating, and sharing knowledge using our Distributed Ontology Gathering Group Integration Environment (DOGGIE) and describe our proof-of-concept experiments. DOGGIE synthesizes agent communication, machine learning, and reasoning for information sharing in the Web domain.

The dynamic behaviour of networks such as business webs is complex and poorly understood. While this is well known, actual studies creating and applying methodologies to quantify and structure the networks both as static snapshots and dynamic, changing landscapes have been relatively few-given the amount of literature devoted to value chains, b-webs and supply/ demand chains.This paper describes a framework for analysis of the structures and dynamics of networked relationships typical in the Internet era. In the paper, a fusion of traditional social network analysis (SNA) methods with business strategies is presented in the context of a wider methodology for strategic network analysis (our "Rosetta Stone"). As a demonstration of this wider concept, a methodology that utilises the results of SNA analysis for determination of company roles in the network of partnerships/ alliances is presented, together with visualisations developed for the Rosetta Stone and a neural network analysis.The case study, examining a real-world network of strategic alliances between 87 companies in the ICT sector, shows in our opinion that the methodology introduced here is capable of capturing the essentials of business networks to provide information for decision makers. Wider applications, e.g. for a new strategic perspective on the mastery of demand/supply networks, are easily identified.

The exploration of social networks is essential for finding capable cooperators who can help problem-solving and for augmenting cooperation between workers in an organization. This paper describes PeCo-Mediator-II to seek capable cooperators through a chain of personal connections (PeCo) in a networked organization. Mo