Vladimir Batagelj | University of Ljubljana (original) (raw)

Uploads

Papers by Vladimir Batagelj

Social network analysis and mining, Jun 25, 2024

Scientometrics, May 28, 2024

Studies in classification, data analysis, and knowledge organization, Dec 31, 2022

arXiv (Cornell University), Mar 1, 2019

In this paper, we present the outer product decomposition of a product of compatible linked netwo... more In this paper, we present the outer product decomposition of a product of compatible linked networks. It provides a foundation for the fractional approach in network analysis. We discuss the standard and Newman's normalization of networks. We propose some alternatives for fractional bibliographic coupling measures.

Springer eBooks, 1996

ABSTRACT

Discrete Mathematics, 2003

Short cycles connectivity is a generalization of ordinary connectivity. Instead by a path (sequen... more Short cycles connectivity is a generalization of ordinary connectivity. Instead by a path (sequence of edges), two vertices have to be connected by a sequence of short cycles, in which two adjacent cycles have at least one common vertex. If all adjacent cycles in the sequence share at least one edge, we talk about edge short cycles connectivity. It is shown that the short cycles connectivity is an equivalence relation on the set of vertices, while the edge short cycles connectivity components determine an equivalence relation on the set of edges. Efficient algorithms for determining equivalence classes are presented. Short cycles connectivity can be extended to directed graphs (cyclic and transitive connectivity). For further generalization we can also consider connectivity by small cliques or other families of graphs.

Cambridge University Press eBooks, Jun 19, 2012

Springer eBooks, Dec 19, 2017

Five key problems of kinship networks are boundedness, cohesion, size and cohesive relinking, typ... more Five key problems of kinship networks are boundedness, cohesion, size and cohesive relinking, types of relations and relinking, and groups or roles. Approaches to solving these problems include formats available for electronic storage of genealogical data and representations of genealogies using graphs. P-graphs represent couples and uncoupled children as vertices, whereas parent-child links are the arcs connecting nodes both within and between different nuclear families. Using results from graph theory, P-graphs are shown to lend themselves to solutions of the problems discussed. Relinking of families through marriage, for example, can be formally defined as sets of bounded groups that are the cohesive cores of kinship networks, with nodes at various distances from such cores. The structure of such cores yields an analytic decomposition of kinship networks and constituent group and role relationships. The Pgraph and Pajek programs for large network analysis help both to represent kinship networks and their patterns and to solve problems of analysis.

Encyclopedia of Social Network Analysis and Mining, 2014

Wiley eBooks, Aug 22, 2014

Scientific citation and other bibliographic networks The literature covering all of the many scie... more Scientific citation and other bibliographic networks The literature covering all of the many scientific disciplines is vast. As a result, its citation network is enormous. It can be viewed as one whole network, or attention can be focused on coherent parts of it. Our strong preference is for the latter strategy. For the substantive reasons laid out in Section 1.3, we pursue only two of these scientific citation networks here. One is for the centrality literature 1 with the other featuring network analysis as a whole. 4.1 The centrality citation network As we note in Section 1.3.1, the centrality literature was started by the Bavelas (1948) paper reporting on experiments for task-oriented groups. The experimental control variable was the communication structure of a task-oriented group charged with distributing information to all of its members. The outcome variable was the time it took the group to complete its task. While centrality did not appear explicitly in this paper, the experimental outcomes can be expressed as: 1) centrality in the network is predictive of actor outcomes, and 2) network centralization is predictive of the task completion times. A research literature was spawned featuring social psychology, small group processes, experimental design, social organization, business administration, and measurement as substantive and technical concerns. Centrality was seen as a useful concept linking these diverse topics, one with great implications for the study of social network processes. Indeed, the above two basic findings morphed into two mantras for social network analysts: 1) actor locations in networks affect actor outcomes, and 2) group network structures matter for collective outcomes. 1 This relatively small part of the network literature, as shown below, has almost a million works and close to two million citation arcs.

Social network analysis and mining, Jun 25, 2024

Scientometrics, May 28, 2024

Studies in classification, data analysis, and knowledge organization, Dec 31, 2022

arXiv (Cornell University), Mar 1, 2019

In this paper, we present the outer product decomposition of a product of compatible linked netwo... more In this paper, we present the outer product decomposition of a product of compatible linked networks. It provides a foundation for the fractional approach in network analysis. We discuss the standard and Newman's normalization of networks. We propose some alternatives for fractional bibliographic coupling measures.

Springer eBooks, 1996

ABSTRACT

Discrete Mathematics, 2003

Short cycles connectivity is a generalization of ordinary connectivity. Instead by a path (sequen... more Short cycles connectivity is a generalization of ordinary connectivity. Instead by a path (sequence of edges), two vertices have to be connected by a sequence of short cycles, in which two adjacent cycles have at least one common vertex. If all adjacent cycles in the sequence share at least one edge, we talk about edge short cycles connectivity. It is shown that the short cycles connectivity is an equivalence relation on the set of vertices, while the edge short cycles connectivity components determine an equivalence relation on the set of edges. Efficient algorithms for determining equivalence classes are presented. Short cycles connectivity can be extended to directed graphs (cyclic and transitive connectivity). For further generalization we can also consider connectivity by small cliques or other families of graphs.

Cambridge University Press eBooks, Jun 19, 2012

Springer eBooks, Dec 19, 2017

Five key problems of kinship networks are boundedness, cohesion, size and cohesive relinking, typ... more Five key problems of kinship networks are boundedness, cohesion, size and cohesive relinking, types of relations and relinking, and groups or roles. Approaches to solving these problems include formats available for electronic storage of genealogical data and representations of genealogies using graphs. P-graphs represent couples and uncoupled children as vertices, whereas parent-child links are the arcs connecting nodes both within and between different nuclear families. Using results from graph theory, P-graphs are shown to lend themselves to solutions of the problems discussed. Relinking of families through marriage, for example, can be formally defined as sets of bounded groups that are the cohesive cores of kinship networks, with nodes at various distances from such cores. The structure of such cores yields an analytic decomposition of kinship networks and constituent group and role relationships. The Pgraph and Pajek programs for large network analysis help both to represent kinship networks and their patterns and to solve problems of analysis.

Encyclopedia of Social Network Analysis and Mining, 2014

Wiley eBooks, Aug 22, 2014

Scientific citation and other bibliographic networks The literature covering all of the many scie... more Scientific citation and other bibliographic networks The literature covering all of the many scientific disciplines is vast. As a result, its citation network is enormous. It can be viewed as one whole network, or attention can be focused on coherent parts of it. Our strong preference is for the latter strategy. For the substantive reasons laid out in Section 1.3, we pursue only two of these scientific citation networks here. One is for the centrality literature 1 with the other featuring network analysis as a whole. 4.1 The centrality citation network As we note in Section 1.3.1, the centrality literature was started by the Bavelas (1948) paper reporting on experiments for task-oriented groups. The experimental control variable was the communication structure of a task-oriented group charged with distributing information to all of its members. The outcome variable was the time it took the group to complete its task. While centrality did not appear explicitly in this paper, the experimental outcomes can be expressed as: 1) centrality in the network is predictive of actor outcomes, and 2) network centralization is predictive of the task completion times. A research literature was spawned featuring social psychology, small group processes, experimental design, social organization, business administration, and measurement as substantive and technical concerns. Centrality was seen as a useful concept linking these diverse topics, one with great implications for the study of social network processes. Indeed, the above two basic findings morphed into two mantras for social network analysts: 1) actor locations in networks affect actor outcomes, and 2) group network structures matter for collective outcomes. 1 This relatively small part of the network literature, as shown below, has almost a million works and close to two million citation arcs.