Conference Program 2007 - Abstracts

Although social network research often focuses on a single relation, multiple relations between the same people can substantially enrich the understanding of the observed networks. The extent to which different types of ties - such as advice and friendship - go together (tie-multiplexity), or the extent to which a tie of one type directed from person i to person j is returned by a tie of another type from j to i (tie-exchange) can provide substantial information about the social processes in a network.
Such mechanisms might especially be relevant when different groups of actors can be distinguished within a network, because it enables one to answer such questions as: “Do some (groups of) actors in the network combine different types of ties more, while others select different actors for a specific type of resource?” Or, “Is there more dyadic exchange between actors of different groups than between actors within the same group?”
Following the logic proposed in Skvoretz and Agneessens (2007), we introduce standardized measures for reciprocity, multiplexity and exchange for actors within the network, and subsequently propose statistical tests for comparing the level of reciprocity, multiplexity and exchange within and between groups of actors in a network. We illustrate the possibilities of these measures and its theoretical relevance with a number of different examples.

In this presentation we consider a number of alternative explanations for the link between interpersonal trust and an actor’s job-satisfaction. Many studies on social networks have argued that informal social relations at work are important predictors of job satisfaction. Social capital researchers stress that social relations are an important source of social support, influence, and information for individuals to reach their goals. Knowing many trustworthy people can help solve problems at work, and is therefore likely to increase an employee’s perceived well-being. Trust relations to specific others can also function as channels of interpersonal influence, in which information, opinions and sentiments are transmitted between the two actors through unconsciously mimicking of, or adjusting to, each other’s expressions. Hence, an individual’s job satisfaction is likely to be influenced by the level of satisfaction of those to whom they have a strong tie.
More recently, scholars have started to point towards two other general processes which focus on the sentiments and attitudes shaping social structure (i.e. social selection processes). First, other actor’s moods and sentiments and similarity can make them more or less attractive as a potential interaction partner. The key idea underlying this argument is that individuals choose their interaction partners based on either the latter’s attributes, or the similarity between both. Moreover, selection might be based on structural characteristics surrounding network of the actors.
We simultaneously consider the above social influence and social selection mechanisms using intra-organizational longitudinal social network data with SIENA. We find no effect of the differences in position in a trust network on job satisfaction (social influence). However, we do find that employees with a high level of job satisfaction are in fact less likely to develop trust in others than people with a lower job satisfaction (social selection). Moreover we do find a tendency between actors to choose others with a similar level of job satisfaction.

Finding suitable analysis techniques for networks generated from social processes is a difficult task when the population changes over time. Traditional social network analysis measures may not work in such circumstances. It is argued that agent-based social networks should not be constrained by a priori assumptions about the evolved network and/or the analysis techniques. In most agent-based social simulation models, the number of agents remains fixed throughout the simulation; this paper considers the case when this does not hold. Thus the aim of this paper is to demonstrate how the network signatures change when the agents' population depends upon endogenous social processes. We argue for a much wider attention from the social simulation community in addressing this open research problem. It seems evident that, in many cases, the size and structure of social networks is contingently constructed by intricate social processes, rather than determined in a rigid and explicit manner. To capture and understand such phenomena it would seem sensible to develop models that embedded such a contingent development of social networks. However this makes the task of choosing suitable techniques for understanding the emergence of networks in such models much more difficult. Scale-independent measures such as nonparametric statistics are useful of in indicating the presence of change in dynamic social networks, leading to further research in the analysis of dynamical networks.

The detection of densely connected subgroups in social networks is of interest in many research fields and several methods have been developed so far to observe these structures. However, many of them are computationally (very) expansive and therefore not applicable for large datasets. Given for example a dataset of user interactions in online platforms, we have to cope with thousands of nodes and edges. For such large graphs it takes most methods an unreasonable amount of time to detect subgroups. The performance problem is even worse if we want to observe the dynamics in these networks. Usually the dataset is split into time-slices and each slice is analyzed separately resulting in an even higher complexity. To deal with this performance problem, usually the data “noise” is removed, thus obtaining a more manageable dataset. However, this process takes extra resources and many thresholds such as minimum degree or edge strength need to be determined to reduce the size of the dataset.

This paper discusses the insights of a field research focused on the interface between communication and psychic system in primary schools’ classrooms, observed as educational networks. The research involved eleven classrooms in nine primary schools in Northern Italy (twelve educators and about 250 pupils, ages 9 from 11). Ninety hours of videotaped educational interactions were produced.
According to social systems’ theory (Luhmann, 1984) we observe education as a communication system that could be empirically analyzed in its basic operations, communications, through the analysis of interactions. We use etnomethodological Conversation Analysis for the analysis of interactions and social network analysis to modelize the forms of educational networks.
This methodology allows us to recognize the forms of structural coupling between educational communication and psychic system, in an environment represented by the classroom as a network of educational relationships.
Educational communication deals with the forms of structural coupling, i.e. the distinctions which make it possible for psychic systems to understand communication, attributing motives and responsibilities for actions, under the influence of expectancies attached to social roles in the educational networks.
Educational networks appears to operate as autopoietic systems (Maturana & Varela, 1980), producing the elements out of which they exist, communications, by means of a network of these elements themselves. They don't import life from their environment: external factors do not directly interfere with the functioning of the system but have to be translated into internal elements.
Our data show that the forms of structural coupling, i.e. the distinctions which make it possible for psychic systems to understand communication, attributing motives and responsibilities for actions (Vanderstraeten, 2000) according with expectancies connected to specialized roles in the educational networks
1) make the outcomes of educational communication unpredictable.
2) blindly re-produce, because of their self-referential operational mode, the forms of educational networks, characterized by low levels of self-reflexivity and openness for structural evolution.

Whilst the literature on interlocking directorships of the corporate world is well-established, less information has been published on such links between political organisations. Research on quangos in the UK has typically taken a demographical rather than a biographical approach, focussing on quantifying the numbers of connections public directors hold rather than assessing the nature of those interlocks. Such work, therefore, has focussed on the overall levels of social capital found in public boardrooms rather than assessing the nature of these individuals who form the UK political elite.

This paper attempts to make use of recent developments in social network analysis to evaluate the dynamics of personnel transfers over the season 2006/2007 and performance of European football teams. We argue that transfers of players heavily depend on the team’s success in the previous year. Teams develop different personnel recruitment strategies according to their position in the league and economic abilities. Furthermore, it is discussed whether network position and personnel transfers relate to the success of a team in the current season.

Research suggests that academic departments’ prestige might be closely related to their location in the web of professor-department hiring relationships. These relationships are the consequence of a series of career choices and opportunities ranging from the prestigious schools graduating more students, to matriculation out of academia, and to upwardly mobile career moves. This paper uses a data set of Sociology Department hiring histories to establish the relationship between prestige and hiring network centrality. In addition it tests whether the relationship is robust to methodological variations and whether the relationship it is independent of factors included in both centrality and prestige measurements.

Appraisal and feedback are key features within any creative process. In design, appraisal allows the designer to progress from one design stage to another. A prototype, for example, can be deemed suitable for development once the design has gone through an appraisal process. Appraisal can occur with novice designers being critiqued by more experienced designers. In an educational context this would be a traditional student/tutor relationship. Appraisal also occurs between peers, an aspect of the design studio which is increasingly encouraged in light of pedagogical rhetoric supporting peer based learning (Nicol/Pilling 2000). This paper seeks to understand this peer appraisal network, and in particular the impact that one’s role within the network has on design work. To do this a mixed methodological approach was undertaken. From the undergraduate design course that was studied, ethnographic type qualitative data was gathered. Additionally, questionnaires were also used, in which the students were asked to evaluate who they sought for feedback. This mixed methodological approach enabled the survey data to reveal the connections that existed in the feedback network, whilst the qualitative descriptions revealed why those connections existed. This paper considers go-between, vulnerable and clique member students and proposes possible reasons why the students in question were certain network roles. This paper also examines how and why go-between students were, on average, producing better design work whilst vulnerable students were not fulfilling their grade potential. It is proposed that go-between students are in a better position to gain feedback from the broadest range of student designers and in so doing are continually improving and refining their work. They are also in a position to impart their own subjective opinion about “good” design work to the design studio as a whole.

The effects of networks on adoption behavior may vary over time in response to institutional environment. We find support for this argument in an analysis of the diffusion of a job search training program among Finnish employment offices. The results indicated that the adoption behavior of collaborative organizations and brokerage position in the local network were related to early adoption of the program. The results further showed that the adoption behavior of structurally equivalent organizations tended to inhibit early adoption of the program. Finally, the results indicated that brokerage position on the national level hindered diffusion.

Social spaces such as blogs, wikis and online social networks are enabling the formation of online communities where people are linked to each other through direct profile connections and also through the content items that they are creating, sharing, tagging, etc. As these spaces become bigger, more intuitive ways of navigating the associated information become necessary. Adding annotations to content items in social spaces (e.g., using topic tags, geographical pinpointing, etc.) is a useful way to aid in the browsing and location of relevant items and of people with similar interests. The Semantic Web aims to link identifiable objects to each other and to textual strings via relationships and attributes respectively. Because data can be gathered from various places and linked together using a common representation format, the Semantic Web is a useful platform for gathering diverse information from heterogeneous social spaces and for performing operations on such linked data. In this paper, we will demonstrate how this linked data, gathered through a semantic search of social spaces, can provide an enhanced view of the activity in a social network. We will explain how, through a further tagging of concepts and disambiguation of terms provided by the Galaxy tool described in this work, objects from social spaces can be further augmented by highlighting related objects, recommending tags, and suggesting relevant sources of knowledge.

Recently, a number of researchers studied complex networks in a variety of domains, such as economics, telecommunications, and biology. The existing approaches to the analysis of such networks are typically built on ideas rooted in graph theory, while network quantitative characterization is usually done in terms of one of the well-established generating mechanism models, such as multiplicative growth, preferential attachment, or behavioral optimization. These approaches, however, can hardly accommodate and comprehensively explain the range of phenomena observed in complex social networks, as 1) few (if any) of the underlying theoretical model assumptions are confirmed with procedures other than distribution fitting and model simulation (i.e. no apparent verification from relevant studies in psychology and sociology), and 2) the presumed generating models are often not properly in accordance with empirical data.

This study aims to develop a reasonably universal theoretical framework that would provide a distinct (from graph theory) analysis perspective and compensate for the above deficiencies.

The proposed theory stipulates a 2-step analysis. First, a social network is considered as a complex system – a conglomerate whose internal mechanisms and structure cannot, for some reason, be inspected in full. A model describing the observed dynamics for one (arbitrary) system state is sought. Using the apparatus of statistical physics, it is shown that a finite mixture of exponentials may often be an appropriate choice. Second, the resulting “one-state” model is generalized by considering the case where its parameters depend on the hidden (i.e. “internal,” not observed) dynamics and properties of the system constituents. A model accounting for the observed “many-states” dynamics can then be formulated as, essentially, linear combination of Laplace-Stieltjes transforms of the distribution functions describing the internal dynamics. The theory is illustrated with an analysis of patent authorship, citation, e-mail, and consumer Web-service networks. It is demonstrated that the suggested framework not only allows for remarkably accurate prediction of the statistics of a network and offers explanations for various “peculiarities” in its behavior (e.g. burstiness), but also provides opportunities for cross-domain checking and validation of the theoretical assumptions. Possible extensions of the theory (e.g. by incorporating ideas from Tsallis thermodynamics) are briefly discussed.

The intensive and increasingly common use of information and communication technologies (ICT) is being extensively studied from different methodological and epistemological perspectives. The dynamic aspects related to concepts, discourses and vocabulary of the virtual environments are discussed on the literature, as well as the power relationships and the networks of sociability and cooperation that are established inside the electronic net which condenses both time and space. However, such analyses are forgetting to include the immigrants who are arriving to western societies from countries that are considered as non-developed, thus ignoring what their discourse about science and technology. Even though access to information is one of the basic Human Rights, in the last couple of decades the digital/informational divide has been rapidly growing. This unequal access to digital information directly and strongly affects social groups whom already have had severe disadvantages in the social, economic, and political areas. Many of these groups are located in Third World (Majority World) countries, but an important number of them are located in developed countries such as Spain, as well. This work proposes a research strategy based on discursive network analysis in order to analyze metaphoric discourse on science and technology (technoscience) by recent sub-saharian immigrants in Spain. With this we hope to inquire on the origin of such metaphors and their perceived utility in the process of adaptation to the Spanish society. A network analysis of discourse about metaphors of technoscience in two periods of the life of the immigrants will be carried out: at the time when they arrive to Spain, and one year after that. This paper shows how data was gathered for analysis for the first part of the study, and it gives a preliminary portrait of technoscience in the context of immigration by using Pajek and Visone.

In the literature, it is argued that because of the shift of activities that once took place in so-called ascribed contexts (family, neighborhood), to activities nowadays taking place in purposive contexts (social and economic organizations), social contexts have become unbundled. Accordingly, people exchanged multiplex relationships for uniplex ones, i.e. one works with one set of people, lives together with another set and spends his or her leisure time with a third set of people. From this perspective, it is important to know to what extent the social structure indeed has become ‘unbundled’.
In this paper, we examine the degree of (un-)bundling by looking at overlaps among a number of social contexts where people meet friends. In a representative survey among Dutch citizens (The Survey of the Social Networks of the Dutch - 2007), we collected detailed information on people’s personal networks. Amongst other things, we asked the respondents where, on which occasion, they got to know their friends, as well as where they continue to meet each other (e.g. at school, at the workplace, in the neighborhood, via family, at a voluntary association, etc.). The obtained information offers the opportunity to study the overlap structure among social contexts, by converting ‘persons by meeting contexts’ matrices to ‘contexts by contexts’ matrices. Questions like ‘Which combinations of social contexts to meet friends are more likely than others?’, ‘Do friends continue to meet each other in other contexts than the context in which they got to know each other?’, and ‘Do people meet their friends in the same or in various contexts?’ can then be answered. Using a number of conceptual dimensions of social contexts (as e.g. can be found in Feld’s ‘focus theory’), we aim at structural explanations for the overlap structure among the contexts in which people meet their friends.

The aim of this research is to develop network strategies that effectively integrate knowledge, in inter organizational R&D networks. ISAPP, the Integrated System Approach for Petroleum Production, is the main case considered. ISAPP is a R&D program of Shell, the Dutch institute for applied research, TNO, and the Technical University Delft. The program aims to improve the recovery of oil fields. Decision makers guide the integration of forty PhD projects into applicable Smart Field solutions. But ISAPP is an almost ‘virtual’ network, crossing the spatial boundaries of five locations. Can collaboration and networking across organizational boundaries be stimulated, such that innovations result? Previous research suggests that openness, mass and focus are properties of knowledge networks that can be adjusted to realize effective results. This research provides ISAPP managers with feedback on the results their network interventions bring about. The challenge is to find the right trade offs between openness, mass and focus, at different stages of program development.

The present study has the aim of exploring the conceptual networks associated to the term “social networks” inside the academic literature on communication and information technologies (ICT). A data base was created form relevant literature from some of the most used data bases containing bibliographic references on topics related to ICT (EBSCO Host, Sage Journal Online, ProQuest, Sociological Abstracts, etc.). A total of 639 records were obtained containing the keywords “social networks” in the context of social studies of ICT. Once data was cleaned of possible repetitions and/or similitudes two symmetric matrices were constructed. The first one contained relationships among keywords, and the second matrix related the names of the main authors of the papers with the most relevant keywords contained in each record. A network analysis using Pajek was conducted, which showed at least three clusters of ICT literature containing the keywords “social networks”, related to aspects of immigration, use of technology by youngsters, and social issues related to technology. An analysis of the evolution of such networks in different years is being conducted.

This paper takes advantage of the possibility of a critical perspective afforded by the social network perspective in analyzing the interaction between financial leader and civil societies in the shaping of specific developmental intervention by donor, to examine micro finance program in East Java. Implemented in 2000s, the micro finance was hailed as important experiment in mainstreaming gender concern in poverty alleviation. It would be building social capital, but given the fact that was a marginal issue and patronized system within civil society in East Java, the incomplete and inaccurate information about the community is used to plan development efforts there is a high tendency to allocate project resources to people who are not in most need of them or the truly poor. The objectives of this paper go beyond reporting on the degree of success of the efforts at mainstreaming gender concerns in the micro finance, though it draws upon many such reports. It will raise a few questions essentially historical in nature on how to develop a framework for more efficient local financial leadership on poverty alleviation program.

In this paper, we present the current version of the APES software application, discuss its conceptualization as well as its features and point out possible applications. The software is now available as a userfriendly java application from our website (www.apes-tool.ch).
The main objective of the APES software is to facilitate the use of social network concepts to social science researchers without much methodological background knowledge. The application allows the researcher to store and graphically display process data (event participation of actors) which can also be rendered as a jpg image and saved in xml. Furthermore, the two-mode process data can be transformed into a one-mode network which are then displayed in the form of a target diagram. Basic indicators comlement the application.
We will demonstrate the use of this tool for comparative case studies in the field of politica science by applying it to two cases from our own research.

Numerous studies highlighted the benefits accruing to firms that move into positions of brokerage by forming ties to otherwise unconnected partners. In particular, spanning structural holes, or the network space between unconnected contacts, was found to confer unique benefits of timely access to non-redundant information as well as control over that information. While existing research also implies that having a large number of nonredundant contacts may be more important in some organizational settings than in others, the dominant – albeit implicit – assumption in the literature is that within the same industrial and social system all firms are equally likely to pursue the benefits of brokerage.
We depart from the assumption of uniformly distributed propensities to engage in brokering behavior and instead investigate why even within the same setting different firms may show vastly different inclinations and abilities to become brokers. We thus explore three key factors that lead some firms more than others to seek and build bridging ties: (i) ability, (ii) motivation, and (iii) opportunity to become a broker. Using this conceptualization, we first suggest that brokerage may emerge from complex interactions between firms’ characteristics and the features of social systems in which they are embedded. Second, we show that the formation of bridging ties is not simply a random process but is, rather, a consequence of purposeful strategic action of select actors who are both willing to and able to become brokers. To test our propositions, we use data on firm partnership formation in the global computer industry in the period 1996-2005.

This paper inquires into two local community mailing lists operating in Israel. The study analyzes the mailing list as a form of “community networks”, exploring their potential effect on community (online and off line) community social capital, as drawn from residents’ level of involvement in local activities and organizations.
Sample was composed of 442 residents, living in two medium sized Israeli urban communities.
Analyzing demographic, attributes and ego networks data (using a name generator technique), we showed that the degree of community residents' civic engagement activities is highly associated with individual's social capital. The relations between individual social capital and civic engagement are mediated by individual political efficacy. Additionally, we demonstrated the selective nature of community electronic list membership. List members have more pronounced stock of social capital. We found that electronic community networks reflect the social structural attributes of residents' relational patterns. Yet, unlike the stipulation put forward by both the utopian and the dys-utopian schools, the Internet does not have an automatic or deterministic effect on community civic engagement. In a "hybrid community", using both offline and online channels to communicate, network infrastructure may enhance existing relational patterns, but neither to replace them nor to abolish their pronounced effect. Internet use is, therefore, embedded with existing social relations.

We examine the impact of network density and Internet use on the quality of social relations among Israeli adolescents. The impact of the Internet on quality of life has recently devoted a great deal of scholarly attention, accompanied by sharp disputes in the academic and intellectual world, typically revolving around two contradictory outlooks. While the "dis-utopian" school customarily expects that individualization, urbanization, and globalization, together with the rapid incorporation of ICT in households are destructive elements to the social fabric, subscribers to the utopian perspective tend to believe that the Internet has contributed to the significant emancipation of individuals and identity groups from constraints of time, space, and from even critical elements of social structure.
The survey, collected in 2001, covers a representative sample of 1,000 households in Israel. We began with a random sample of 60 localities with a population of 2,000 or more. Next, we selected neighborhoods, according to the size of the adolescent population in each settlement. The number of neighborhoods in each settlement was determined by the size of the juvenile population (13-18 years old) in the settlement. At least one neighborhood was randomly selected in settlements with a low proportion of adolescents, and more than one in larger urban areas. In each neighborhood we randomly selected 15 households. The selected neighborhoods represent all geographic areas in Israel and different sizes of settlements, from big cities to small towns and villages. The survey includes items on social and demographic characteristics of the youths, socio-demographic characteristics of their closest friends, channels of communication, types of resources exchanged, and the extent of perceived closeness to each friend. The survey asked each adolescent for the names of six close friends. Respondent provided information about each friend’s age, gender, place of residence; whether they met the friend for the first time at school, in the course of extracurricular activities, in the neighborhood, or online. The adolescents were also asked to indicate the length of time they have known the friend, the extent of closeness and trust they felt toward the friend, and whether they would ask for help from each of the friends listed.
The surveys provide information about the amount (average hours) spent on the Internet, the frequency of use, and the type of Internet use interviewees used. The survey also included an open-ended question asking adolescents to indicate the average number of daily hours they were connected to the Internet.
We show that individual social capital and network density are inversely associated. This is a surprising finding for proponents of the social integration perspective. By contrast, the analyses produced corroborating results to the theory of structural holes (i.e. Granovetter, 1973; Burt, 1992), but with an important special twist: controlling for nationality and gender, it seems that Israeli adolescents who know wider kinds of contacts (whether off line or online), and whose alters do not know one another, are able to promote a higher degree of social capital. The importance of these finding for theory of social capital and the study of adolescents’ social networks are discussed.

Sophia-Antipolis is a business park in the south of France, originating from the private initiative of a single man in the early 1970s. Whereas the first two decades of growth in terms of number of firms and employment was merely exogenous (existing international companies relocating to Sophia- Antipolis), the growth has been strongly endogenous (spin-offs and high-tech start-ups) since the early 1990s. We argue this change should have strong implications for the network of inter-firm cooperation in which the firms of Sophia-Antipolis are involved. Particularly, we hypothesize that the changing growth regime from mainly exogenous to highly endogenous has provoked the emergence of local collective learning. That is to say, we expect Sophia-Antipolis to have turned from a satellite platform (with international firms not cooperating locally) into a cluster in which local collective learning practices play a major role. We test this hypothesis by reconstructing the evolution of the inter-firm knowledge transfer network of Sophipolitan firms on the basis of EPO patent data from 1978 to 2002. Taking all patents with at least one inventor from Alpes-Maritimes (the region surrounding Sophia- Antipolis), we suppose knowledge transfer between firms has taken place in case of (1) co-patenting (if two or more firms have co-developed a patent) and in case of multiple applicant inventorship (if an inventor has been involved in the development of patents of multiple firms). The reconstructed networks show a substantial increase in local collective learning, though not at the expense of cooperation at higher spatial scales.

One of the currently recognised strong trends in the economic development literature is that firms align and collaborate in value-added activities, seeking efficiency from specialisation and from capturing synergies and complementarities based on sharing resources horizontally and vertically. Active cooperation between firms generates inter-firm relations that settle in a variety of regional and industrial agglomerations and concentrations that are labelled as clusters.

This paper reports on the design and application of an original multi-stage cluster methodology that utilises multiple statistical methods for cluster analysis and network analysis, and complements them with a detailed analytical procedure for mapping cluster membership and analysis of intra-cluster and inter-cluster relations. The multi-stage cluster methodology was applied for the clustering of regional economic activities related only to the health technology cluster.

The value-chain cluster map reveals a rich set of inter-related activities that feed into each other. Each cluster group contains firms that have similar profile of industry activities. The richness of activities exhibited in this diagram reveal complex value-added activities within the health-care system and the health technology cluster.

Each cluster group was subjected to network analysis which revealed the structure of inter-related industries and firms. A selection of individual cluster maps show the difference between the two R&D cluster groups (‘R&D Generic’ and ‘Medical & Bio-pharma R&D’) and give additional insides into the inter-industry and inter-firm competitive relationships.

Research reveals also that a significant number of clusters are interconnected by a number of industries – as an evidence of convergence and synergies across value chains that companies have already identified and captured within their portfolio of operations.

The results from the cluster analysis and network mapping demonstrated the convergence across the healthcare, instrumentation industry, pharmaceuticals, and bio and medical R&D.

Relational approaches to organization typically ignore processes of network formation while studying network consequences, and, symmetrically, ignore the implications of network structures and flows while studying network antecedents. In this paper we address these symmetric weaknesses in the organizational literature by specifying dynamic models for the joint representation of network selection and influence processes.
We estimate our models using a set of longitudinal data that we have collected on a sample of graduate students enrolled in a professional management degree program, followed over the program’s 17 month period. The networks studied are communication, advice seeking, professional esteem, and friendship. Controlling for a variety of individual sources of partner choice and performance (e.g., age and gender), mutual dependence of the networks on each other, and a number of endogenous local network effects (e.g., transitivity, reciprocity, or preferential attachment), we find that performance of individual subjects tends to be assimilated to the performance of their network neighbours, and that performance also acts as a basis for network formation in some, but not all of the studied networks.

We present a model of a trust-based recommendation system on a social network. The idea of the model is that agents use their social network to reach information and their trust relationships to filter it. We investigate how the dynamics of trust among agents affect the performance of the system by comparing it to a frequency-based recommendation system. Furthermore, we identify the impact of network density, preference heterogeneity among agents, and knowledge sparseness to be crucial factors for the performance of the system. The system self-organises in a state with performance near to the optimum; the performance on the global level is an emergent property of the system, achieved without explicit coordination from the local interactions of agents.

With the migrant tide in China (153 Million internal migrants according to the national statistic data in 2006), more and more young rural people go to city for changing their life. In this paper, we present an ethnography study on how Chinese rural-migrants adapt themselves to the city life in a totally new environment. Reconstructing social network plays an important role during this process.

Granovetter’s theory of “ the strength of weak tie” has great interpretation power in western countries, but it has new implication in China. “Trust ” instead of “information” is the biggest issue for most Chinese rural-migrants seeking jobs in cities. The relationship based on the same hometown or consanguinity are the two most important factors, they can bring “strong trust” even when the two rural migrants had little contact before they came to city.

New technology is also becoming more and more common and they play an important role in Chinese rural-migrants’ city life. In this paper we also explore the role and implication of new technology for rural-migrants in reconstructing their social networks in cities.