Cohesive Subgroups Research Articles

Fast determination of structurally cohesive subgroups in large networks

Structurally cohesive subgroups are a powerful and mathematically rigorous way to characterize network robustness. Their strength lies in the ability to detect strong connections among vertices that not only have no neighbors in common, but that may be distantly separated in the graph. Unfortunately, identifying cohesive subgroups is a computationally intensive problem, which has limited empirical assessments of cohesion to relatively small graphs of at most a few thousand vertices. We describe here an approach that exploits the properties of cliques, k-cores and vertex separators to iteratively reduce the complexity of the graph to the point where standard algorithms can be used to complete the analysis. As a proof of principle, we apply our method to the cohesion analysis of a 29,462-vertex biconnected component extracted from a 128,151-vertex co-authorship data set.

Actor non-response in valued social networks: The impact of different non-response treatments on the stability of blockmodels

Social network data usually contain different types of errors. One of them is missing data due to actor non-response. This can seriously jeopardize the results of analyses if not appropriately treated. The impact of missing data may be more severe in valued networks where not only the presence of a tie is recorded, but also its magnitude or strength. Blockmodeling is a technique for delineating network structure. We focus on an indirect approach suitable for valued networks. Little is known about the sensitivity of valued networks to different types of measurement errors. As it is reasonable to expect that blockmodeling, with its positional outcomes, could be vulnerable to the presence of non-respondents, such errors require treatment. We examine the impacts of seven actor non-response treatments on the positions obtained when indirect blockmodeling is used. The start point for our simulation are networks whose structure is known. Three structures were considered: cohesive subgroups, core-periphery, and hierarchy. The results show that the number of non-respondents, the type of underlying blockmodel structure, and the employed treatment all have an impact on the determined partitions of actors in complex ways. Recommendations for best practices are provided.

Identifying candidate genes for wood formation in poplar based on microarray network analysis and graph theory

In poplar, genetic research on wood properties is very important for the improvement of wood quality. Studies of wood formation genes at each developmental stage using modern biotechnology have often been limited to several genes or gene families. Because of the complex regulatory network involved in the co-expression and interactions of thousands of genes, however, the genetic mechanisms of wood formation must be surveyed on a genome-wide scale. In this study, we identified wood formation-related genes using a differentially co-expressed (DCE) gene subset approach based on biological networks inferred from microarray data. Gene co-expression networks in leaf, root, and wood tissues were first constructed and topologically analyzed using microarray data collected from the Gene Expression Omnibus. The DCE gene modules in wood-forming tissue were then detected based on graph theory, which was followed by gene ontology (GO) enrichment analysis and GO annotation of probe sets. Finally, 72 probe sets were identified in the largest cohesive subgroup of the DCE gene network in wood tissue, with most of the probe sets associated with wood formation-related biological processes and GO cellular component categories. The approach described in this paper provides an effective strategy to identify wood formation genes in poplar and should contribute to the better understanding of the genetic and molecular mechanisms underlying wood properties in trees.

Detecting large cohesive subgroups with high clustering coefficients in social networks

Clique relaxations are used in classical models of cohesive subgroups in social network analysis. Clustering coefficient was introduced more recently as a structural feature characterizing small-world networks. Noting that cohesive subgroups tend to have high clustering coefficients, this paper introduces a new clique relaxation, α-cluster, defined by enforcing a lower bound α on the clustering coefficient in the corresponding induced subgraph. Two variations of the clustering coefficient are considered, namely, the local and global clustering coefficient. Certain structural properties of α-clusters are analyzed and mathematical optimization models for determining α-clusters of the largest size in a network are developed and validated using several real-life social networks. In addition, a network clustering algorithm based on local α-clusters is proposed and successfully tested.

The analysis of criminal and terrorist organisations as social network structures:a quasi-experimental study

The need to tackle organised crime and terrorism threats indicates, among other things, the importance of analysing efficiently and effectively the patterns of criminal ties. Social network analysis (SNA) offers conceptual frameworks, techniques and software tools to measure and visualise empirical networks of any kind. The aim of this study was to test, examine and consider the usefulness of SNA as a principal approach for the analysis and investigation of criminal and terrorist organisations. Two quasi-experiments were deployed, using a conceptual analytical framework consisting of topological, cohesive subgroups, centrality and key player measures, to see whether the application of network concepts and techniques can help law enforcement and intelligence agencies understand and derive meaning from complex and/or large data sets. The results confirmed the research hypothesis and key conclusions were drawn. It was found that SNA is a quite effective and reliable approach that gives both investigators and analysts the power to know criminal and terrorist network structures. Clearly, SNA can influence a new way of thinking and improve professional practice through the development of intelligence-led applications. The study furthers understanding regarding the use of network concepts and techniques for crime analysis and investigation, and it can be used to inform decisions about SNA programmes and technologies.

미래를 향하는 한국 마이크로 패키징 학회지의 과거와 현재 연구영역에 관한 연구

Department of Material Science and Engineering, Yonsei University, 50, Yonsei-ro, Sedaemungu, Seoul 120-749, Korea(Received August 31, 2015: Corrected September 3, 2015: Accepted September 24, 2015)Abstract: After its inception in 1994, the Journal of the Microelectronics and Packaging Society has continued to makesignificant strides in the number and quality of publications within its field. The interest in the microelectronics andpackaging research has become more critical as consumer electronic products continue its increasing trend towards thinnerand lighter devices that tests the boundaries of electronic devices. This study utilizes social network analysis of allpublished literature in the Journal for the past 22 years. Using the keywords and abstracts available within each individualarticle, the publications within the Journal has focused on major topics covering ① flip chip, ② reliability, ③ Cu, ④IMC (intermetallic compounds), and ⑤ thin film. Using the social network relationship between keywords within articles,flip chip was closely associated with reliability, BGA (ball grid array), contact resistance, electromigration in many ofthe published research works within the Journal. From the centrality analysis, it was found that flip chip, reliability, Cu,thin film, IMC, and RF (radio frequency) to have a high degree of centrality suggesting these key areas of research haverelatively high connectivity with other research topics within the Journal and is central to many of the research fieldswithin the micro-electronics and packaging area. The cohesiveness analysis showed research clustering of five majorcohesive sub-groups and was mapped to better understand the major area of research within this field. Research withinthe field of micro-electronics and packaging converges many disciplines of science and engineering. The continuedevolution within this field requires an understanding of the rapidly changing industry environment and the consumer needs.Keywords: social network analysis, keyword co-occurrence, centrality, cohesive subgroup

Exploring network structures via blockmodeling of the shortest path length matrix

Among the various methods of network analyses, blockmodeling is one of the key approaches to analyzing network structures. It can be applied not only to position analysis of the individual nodes, but also to cohesive subgroup discovery of the whole network. In this paper, we explore network structures via blockmodeling of the shortest path length matrix. First of all, we dichotomize the shortest path length matrix with different thresholds to generate reachability matrices within different steps. Then, through blockmodeling of the reachability matrix within n steps, we can analyze the network based on the cohesive subgroup, n-clique. Furthermore, through blockmodeling of the maximum reachability matrix, we can fit the network properly into the bowtie structure.

Affiliative structures and social competence in Portuguese preschool children.

The goal of this study was to determine whether peer social competence (SC), defined as the capacity to use behavioral, cognitive, and emotional resources in the service of achieving personal goals within preschool peer groups, was related to the type of affiliative subgroups to which children belonged. Two hundred forty Portuguese preschool children (152 seen in consecutive years of data collection) from middle-class families participated. Affiliative subgroup type was assessed from observed proximity data. Social competence was assessed using observational and sociometric measures. Children in more cohesive affiliative subgroups had higher levels of SC, whereas ungrouped children had the lowest SC scores. Follow-up analyses indicated that 2 of the measured SC domains (social engagement/motivation, profiles of behavior/personality attributes) were responsible for the overall difference in SC. Further, membership in a more cohesive subgroup in 1 year contributed to increases in scores for 2 of 3 SC domains (i.e., profiles of behavior/personality attributes and peer acceptance) in the following year. Results suggest that affiliative subgroups both reflect and support individual differences in peer SC during early childhood.

Cohesive subgroups in academic networks: unveiling clique integration of top-level female and male researchers

Social networks are said to have a positive impact on scientific development. Conventionally, it is argued that female and male researchers differ in access to and participation in networks and hence experience unequal career opportunities. Due to limited capacities of time and resources as well as homophily, top-level scientists may structure their contacts to reduce problems of complexity and uncertainty. The outcomes of the structuring can be cohesive subgroups within networks of relation. Women in science might suffer exclusion from cliques because of being dissimilar in the arena. The present paper aims to explore integration in and composition of scientific cliques. A three-step analysis is conducted: Firstly, cliques are identified. Secondly, overlap structures are examined. Thirdly, group compositions are analysed in terms of other personal attributes of the researchers involved. Building on network data of female and male investigators, the article applies a comparative case study design including two cutting edge research institutions from the German Excellence Initiative. The study contrasts a Cluster of Excellence with a Graduate School and the corresponding formal with the informal networks. The results imply that the general hypothesis of unfavourably embedded female researchers cannot be supported. Although women are less integrated in scientific cliques, the majority is involved in an inner social circle which enables access to career-relevant network resources.

Trajectories of male sexual aggression from adolescence through college: A latent class growth analysis.

Approximately 25% of male college students report engaging in some form of sexual coercion by the end of their fourth year of college. White and Smith (2004) found that negative childhood experiences-childhood sexual abuse, childhood physical abuse, and witnessing domestic violence-predicted sexual aggression perpetrated before college, but not during the subsequent college years, a puzzling finding in view of the reasonably consistent rates of sexual aggression from adolescence to the first 2 years of college. The current study takes a person-centered approach to sexual aggression in an attempt to resolve this discrepancy. We examined the possibility of cohesive subgroups of men in terms of their frequency of sexual aggression across the pre-college and college years. A series of latent class growth models were fit to an existing longitudinal dataset of sexual experiences collected across four time points-pre-college through year 3 of college. A four-trajectory model fit the data well, exhibiting significantly better fit than a three-trajectory model. The four trajectories are interpreted as men who perpetrate sexual aggression at (1) low (71.5% of the sample), (2) moderate (21.2%), (3) decreasing (4.2%), and (4) increasing (3.1%) frequencies across time. Negative childhood experiences predicted membership of the decreasing trajectory, relative to the low trajectory, but did not predict membership of the increasing trajectory, explaining the discrepancy uncovered by White and Smith. Implications for primary prevention of sexual aggression are discussed.

Friends First? The Peer Network Origins of Adolescent Dating.

The proximity of dating partners in peer friendship networks has important implications for the diffusion of health-risk behaviors and adolescent social development. We derive two competing hypotheses for the friendship-romance association. The first predicts that daters are proximally positioned in friendship networks prior to dating and that opposite-gender friends are likely to transition to dating. The second predicts that dating typically crosses group boundaries and opposite-gender friends are unlikely to later date. We test these hypotheses with longitudinal friendship data for 626 9th grade PROSPER heterosexual dating couples. Results primarily support the second hypothesis: romantic partners are unlikely to be friends in the previous year or share the same cohesive subgroup, and opposite-gender friends are unlikely to transition into dating.

中班幼儿社会交往能力的社会网络分析

幼儿人际交往能力的高低对幼儿的未来发展、适应社会具有十分重要的作用。培养幼儿的人际交往能力对于广大的幼儿教师来说是不可推卸的责任。本研究采用社会网络分析方法对雅安市天全县某幼儿园一中班幼儿的人际交往关系进行了分析，结果表明：1)从网络中心度来看，幼儿关系网络中心度的三个指标(点度中心度、中介中心度和接近中心度)与教师的性格主观评价基本相关，但是不同的指标之间也不完全一致，它们从不同方面反映了幼儿的交往能力情况；2)从威望特征来看，三个反映威望的指标(入度、入域和近距威望)也不完全一致，同样，这些差异也反映出幼儿交往情况的不同方面；3)从凝聚子群特征来看，该班级的交往网络中小团体特征不明显。总之，本研究采用社会网络分析方法对幼儿的社会交往能力进行研究，利用量化的数据指标，有助于比较客观的、多方面的衡量个体幼儿的交往情况，我们可以根据分析所得结果，采取相应的教育策略，有针对性的培养幼儿的人际交往能力。 Children’s interpersonal skills play a very important role in their future development and social adaptation. Cultivating children’s interpersonal skills is the unshirkable responsibility for the preschool teachers. This study adopted the methods of social network analysis to analyze the in-terpersonal relationship in a middle class of a kindergarten in Tianquan county of Ya’an city. The results showed that: 1) from the perspective of centrality, three indicators (indegree, betweenness centrality and closeness centrality) were correlated with the subjective evaluation of the children’s interpersonal skills by the teacher. However, because the different indicators reflect the different aspects of children’s communication ability, they are not entirely consistent; 2) in view of the prestige, the three indexes (popularity, input domian and proximity prestige) are also not entirely consistent. And likewise, the differences also reflect different aspects of interactions between the children; 3) from the point of cohesive subgroups, the characteristic of small subgroup is not obvious. In conclusion, this study adopted the methods of social network analysis to investigate the social intercourse ability of the children, resulting in some quantitative metrics which might be used to let us take corresponding educational strategy and cultivate children’s interpersonal communication ability.

A social network analysis of teaching and research collaboration in a teachers' virtual learning community

AbstractAnalysing the structure of a social network can help us understand the key factors influencing interaction and collaboration in a virtual learning community (VLC). Here, we describe the mechanisms used in social network analysis (SNA) to analyse the social network structure of a VLC for teachers and discuss the relationship between face‐to‐face and online collaborations. In contrast to previous research applying SNA to analyse measuring indexes alone, we emphasise the mechanisms combining SNA, questionnaires, content analysis and focus group interviews—the key methodology to analyse complex interaction in a VLC. On this basis, we present an analysis model for teachers' VLC and apply it to a teachers' VLC known as ‘IRIS’. The study participants comprised 172 K12 teachers aged between 25 and 55 years. This study collected collaboration data from 2006 to 2012 and analysed the social network structure using sociograms, centrality, cohesive subgroups, clique phenomenon, and matrix correlation of SNA. These findings suggest that face‐to‐face and online collaborations are both indispensable in teaching and in research and continuously supplement and remedy each other in professional development. Moreover, the model succeeded in accessing, describing and analysing the social network structure of a VLC.

Editorial board interlocks in financial economics

We examine the patterns of editorial collaboration in 20 leading finance journals, from 1994 to 2013. We investigate the social structure of editorship across journals, editorial positions, nationalities and affiliations. Our sociological argument is structured upon Bourdieu's (1988) concept of social and symbolic capital and Whitley's (2000) concept of norm. Our analysis identifies elite scientists who influence the orientation as well as the content of published research in finance. We find that the academic community of editors is rather permeably structured; it is spread around a dense core and a relatively scattered periphery. Moreover, we find that editorial board membership has become increasingly accessible to scientists of diverse nationalities, while, within the network, the number of cohesive subgroups decreases across the sample period.

Looking Back at Steel Research International and Its Future

The knowledge structure of steel research in the journal of Steel Research International from 1990 to 2013 was identified by extracting and analyzing 5700 keywords from 2430 papers in the Journal. Analysis of the keyword frequency over a 24‐year period (1990–2013) identified keywords that consistently dominated the field of steel research including microstructure, finite element method, transformation induced plasticity, mechanical properties, continuous casting. By incorporating social network analysis to the identified keywords, a high degree of centrality was microstructure, finite element method, continuous casting, and slag. From the co‐occurrence matrix of cohesiveness, eight cohesive subgroups were identified. Additionally, subgroup mapping reflected research development trends in each of the eight subgroup research areas. By incorporating the author's country of origin the frequency of articles published by country was ascertained and found to be in the order of Germany, China, Sweden, Japan, South Korea, India, Austria, and USA. By identifying the trends of steel research, confirmed sub‐research areas, and revealing core keywords, the results of this study provide valuable insights into understanding the knowledge structure of the steel research field and research directions for Steel Research International.

Security Visualization Analytics Model in Online Social Networks Using Data Mining and Graph-based Structure Algorithms

The rise of the Internet accelerates the creation of various large-scale online social networks, which can be described the relationships and activities between human beings. The online social networks relationships in real world are too big to present with useful information to identify the criminal or cyber-attacks. This research proposed new information security analytic model for online social networks, which called Security Visualization Analytics (SVA) Model. SVA Model used the set of algorithms (1) Graph-based Structure algorithm to analyze the key factors of influencing nodes about density, centrality and the cohesive subgroup to identify the influencing nodes of anomaly and attack patterns (2) Supervised Learning with oneR classification algorithm was used to predict new links from such influencing nodes in online social networks on discovering surprising links in the existing ones of influencing nodes, which nodes in online social networks will be linked next from the attacked influencing nodes to monitor the risk. The results showed 42 influencing nodes of anomaly and attack patterns and can be predict 31 new links from such nodes were achieved by SVA Model with the accuracy of confidence level 95.0%. The new proposed model and results illustrated SVA Model was significance analysis. Such understanding can lead to efficient implementation of tools to links prediction in online social networks. They could be applied as a guide to further investigate of social networks behavior to improve the security model and notify the risk, computer viruses or cyber-attacks for online social networks in advance.

Cohesive subgroups in the international collaboration network in astronomy and astrophysics

This study employs social network analysis to identify institutions with strong international collaborative relationships in astronomical research. We find that the strongest ties tend to link institutions across continents in research collaboration. However, the effect of geographic factors is still notable in light of the fact that most of the institutions in the largest subgroup are located in Europe. Examination of the network position, measured by degree centrality, indicates that homophily is more common than heterophily in the network. A relatively high number of relational ties are observed among institutions that have similar levels of network centrality. Mutual relations are prevalent among central institutions, while strong mutual solidarity exists between institutions on the periphery of the network. This study shows a general unstable international collaborative relationship among astronomical institutions. While more and more institutions have linked up in research collaboration, many of them keep relatively weak ties. Institutions tend not remain in the same subgroup, but link to different partners over time.

Exact algorithms for problems related to the densest k-set problem

Many graph concepts such as cliques, k-clubs, and k-plexes are used to define cohesive subgroups in a social network. The concept of densest k-set is one of them. A densest k-set in an undirected graph G=(V,E) is a vertex set S⊆V of size k such that the number of edges in the subgraph of G induced by S is maximum among all subgraphs of G induced by vertex sets of size k. One can obtain a densest k-set of G in O(k2nk) time by exhaustive-search technique for an undirected graph of n vertices and a number k<n. However, if the value of k approaches n/2, the running time of the exhaustive-search algorithm is O⁎(2n). Whether there exists an O⁎(cn)-time algorithm with the fixed constant c<2 to find a densest k-set in an undirected graph of n vertices remains open in the literature. In this paper, we point out that the densest k-set problem and a class of problems related to the concept of densest k-sets can be solved in time O⁎(1.7315n).

Networks containing negative ties

Social network analysts have often collected data on negative relations such as dislike, avoidance, and conflict. Most often, the ties are analyzed in such a way that the fact that they are negative is of no consequence. For example, they have often been used in blockmodeling analyses where many different kinds of ties are used together and all ties are treated the same, regardless of meaning. However, sometimes we may wish to apply other network analysis concepts, such as centrality or cohesive subgroups. The question arises whether all extant techniques are applicable to negative tie data. In this paper, we consider in a systematic way which standard techniques are applicable to negative ties and what changes in interpretation have to be made because of the nature of the ties. We also introduce some new techniques specifically designed for negative ties. Finally we show how one of these techniques for centrality can be extended to networks with both positive and negative ties to give a new centrality measure (PN centrality) that is applicable to directed valued data with both positive and negative ties.

Nonprofit-Business Partnering Dynamics in the Energy Efficiency Field

What explains differentiation in a nonprofit’s organizational practices around partnering with businesses? I propose that attendance at events plays a role. To explore this, I identify a network of energy and environmental nonprofit organizations and the events that brought them together to influence appliance energy efficiency from 1994 to 2006. Using network analysis to identify cohesive subgroups, I find that over time, organizations become more likely to choose one type of event over another suggesting niche development occurred in the field. I also find that, controlling for previous efforts with businesses, funding, and mission, organizations that belonged to a cohesive subgroup of organizations brought together by an annual event promoting cooperative market approaches from 2001 to 2006 were five times more likely than those nonprofits in other subgroups to partner with business. This research has implications for understanding the creation of new events and its impact on organizational practices.