PROCESS-ORIENTED COORDINATION OF

COLLABORATIONS IN SOCIAL NETWORKS

Agnes Koschmider, Andreas Oberweis and Huayu Zhang

Institute of Applied Informatics and Formal Description Methods (AIFB), Karlsruhe Institute of Technology (KIT)

Kaiserstrasse 89, 76133 Karlsruhe, Germany

Keywords: Processes, Collaboration, Community, Social Networks, Modeling.

Abstract: Social networks are known to stimulate the exchange and sharing of information among peers. Even more

social networks can initiate a cooperation (e.g., people sharing music) and a collaboration (e.g., searching

for collaborators for research works). However, social networks are not widely used as work resources (e.g.,

for help or support request) mostly due to missing coordination mechanisms. This paper describes how col-

laboration can be coordinated in social networks. The proposed way to achieve this is based on the usage of

a set of activity lists of social network members. An activity list specifies all personal activities required to

reach a collaborative output. Based on the activity lists a process model can be generated that controls and

analyzes the coordination. Activities requiring collaboration are performed using social network. The ap-

proach is illustrated with a use case.

1 INTRODUCTION

Social networks such as Facebook, MySpace, Lin-

kedIn and XING have attracted millions of users

over the past few years. The platforms are mainly

used for private purposes such as the initiation and

maintenance of personal contacts, but also for busi-

ness purposes, for example, job placement or mar-

keting.

Although there are functionalities available in

social networks for (research) collaboration (e.g.,

entry of interests), social networks are not widely

used as work resources (e.g., for help or support

requests). One reason for this is an insufficient coor-

dination support for collaborations. Ineffective

communication mechanisms in social networks also

hamper the activation of collaboration (e.g., there is

no support for analysis of interpersonal relation-

ships). Particularly, there is a lack of coordination of

geographically distributed collaboration among dif-

ferent organizations (e.g., writing an EU proposal)

where geographical, language or technical barriers

exist. Coordination mechanisms can be used to ana-

lyze the existing relationships, help to overcome

communication barriers regarding the output of col-

laboration and may efficiently organize collaborative

activities.

Let the following scenario be given as shown in

Figure 1. Member B1 from the social network on the

right hand side intends to write a research paper

(preferably with people of complementary know-

ledge). New in the department, she is not aware of

the paper writing process. Assume a source is given

that describes the writing process (e.g., a wiki page)

or the user specifies by herself the activities required

to write the paper. Based on this source, respectively

activity list, a process model is generated including

her own activities. Some of these activities might

require collaborators who can be found in social

networks, for example, the collaborators A3 and B9.

Subsequently, the initial process model of B1 can be

extended with activities of these two collaborators.

Figure 1: Coordinating the paper writing process.

361

Koschmider A., Oberweis A. and Zhang H.

PROCESS-ORIENTED COORDINATION OF COLLABORATIONS IN SOCIAL NETWORKS.

DOI: 10.5220/0002855703610366

In Proceedings of the 6th International Conference on Web Information Systems and Technology (WEBIST 2010), page

ISBN: 978-989-674-025-2

This paper describes a model for the coordina-

tion of collaborations in social networks. Particular-

ly, the focus lies on geographically distributed colla-

borations, which are difficult to handle. The ap-

proach presented in this paper can be used to en-

hance existing social networks supporting the coor-

dination of (research) collaborations.

The paper is structured as follows. Section 2 de-

scribes the properties of geographically distributed

research collaborations in social networks and the

activities to be coordinated. Section 3 provides a

process-oriented approach to coordinate such colla-

borations. The modeling process is explained in Sec-

tion 4. The approach is illustrated with a use case in

Section 5. Related work is discussed in Section 6.

The paper concludes with a summary and an outlook

on future research in Section 7.

2 RESEARCH COLLABORATION

IN SOCIAL NETWORKS

2.1 Fundamentals of Social Networks

A social network is a network whose nodes are so-

cial actors (individuals or groups) and whose edges

represent the relationships of actors to each other

(Barnes, 1972). This paper will refer to research

collaborations in online social networks where social

actors (network members) are research groups,

scientists, technical staff, doctoral or postdoctoral

researchers. Each edge is weighted with a positive

value, which indicates the frequency of communica-

tion between two network members. Collaboration

starts in a social network as soon as the first two

stages of network development (Potential and Coa-

lescing, see Figure 2) are completed. After reaching

the end of the third stage (Active) the network mem-

bers stay only occasionally in contact. Figure 2 also

provides an overview about the activities of each

stage, to which we will refer in the following sec-

tions.

2.2 Properties of Geographically

Distributed Research

Collaborations

A collaboration can be described with the following

four properties (Schramm-Klein, 2005): direction of

collaboration (vertical, horizontal, lateral), number

of partners (bilateral, multilateral), formalization of

collaboration ("strong ties", "weak ties") and bind-

ing intensity (formal, informal collaboration). In

Figure 2: Stages of social network development (Wenger,

1998).

addition to these properties, a research collaboration

can be characterized by the type of collaboration:

 Reading and writing: The collaboration artefact

is a result of a knowledge creation process, such

as a model, a project proposal or program code.

 Organizational: The collaboration artefact is a

result of an organizational process, such as a

joint workshop or a business trip.

In this paper a geographically distributed re-

search collaboration in social networks is considered

as a not by contract regulated teamwork performed

on the two types of collaboration. We assume that

two or more network members are involved in the

collaboration, who have weak relationships with

each other. These network members have either the

same or complementary research interests.

To support a collaboration according to the so-

cial network development, different activities in the

first three development stages should be coordinated

with respect to the above-mentioned properties. Ta-

ble 1 summarizes the corresponding activities.

Table 1: Summary of activities and collaboration proper-

ties for each development stage.

Stage Activity to be carried out Property

Potential Finding partners horizontal

vertical

bilateral

multilateral

Coalescing Building relationship (estab-

lishment of contacts and

communication), specifying

collaboration outputs

bilateral

multilateral

informal

„weak ties“

Active Executing collaboration

(e.g. writing a publication,

joint assessment of a pro-

posal, development of a

prototype and organisation

of a joint workshop)

bilateral

multilateral

informal

„weak ties“

reading and

writing

organizational

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

362

3 MODEL FOR COORDINATION

OF COLLABORATION IN

SOCIAL NETWORKS

To coordinate a research collaboration in social net-

works a process-oriented approach has been chosen.

The advantage of this approach is a controlled coor-

dination and analysis of the collaboration. In the

following we will define a model that includes all

activities of social network members in order to

reach a collaboration output. The model supports the

consultation of a social network in case of a collabo-

ration. The process model that is generated based on

the activity list of a social network member is called

a “Community Process”.

The Community Process (CP) is a set of related

activities of network members that are executed to

achieve a collaboration output. The activities of a

Community Process are either Single Activities or

Collaborative Activities. In a Single Activity only

one or no network member is involved. In a Colla-

borative Activity at least two network members are

involved that are connected based on an explicit

collaborative relationship. Social networks are used

to implement Collaborative Activities, which de-

scribe the involvement of network members in social

production (Benkler, 2006). The activities are per-

formed sequentially, in parallel, iteratively or alter-

natively. Each Community Process has exactly one

start and one end activity. A Community Process can

be decomposed into several sub-processes and has at

least one Collaborative Activity that refines a se-

quential sequence of sub-processes Finding Part-

ners, Building Relationship and Collaboration Ex-

ecution. Examples of a Community Process are "col-

laboration in an EU proposal", " collaboration in the

organization of a workshop" and "idea generation

process". Another example of a Community Process

will be discussed in detail in Section 5.

A Community Process is associated with a set of

process resources that are designated as Community

Process Objects. A Community Process Object is

either a Flowing Object or a Non-flowing Object.

The Flowing Object includes all information and

data that will be transferred from one activity to

another so that an activity can be performed. The

Non-flowing Object includes those resources that

execute the activity.

A special Non-flowing Object is Community Us-

er (CU) that describes exactly one network member

through a user profile. A Community User has rela-

tionships to other Community Users. From these

relationships the network structure can be derived. A

user profile contains information about e.g., contact

information, knowledge, experience and interests. A

friendship and/or a relationship of knowledge may

exist between two Community Users.

A type of Flowing Object is Community Content

(CC) that is a container for a time, place or event-

limited context. Examples of Community Content

for research collaborations are "publication", "re-

search project application", "workshop" and "confe-

rence".

The collaboration will be coordinated by specify-

ing a Community Process of a network member,

especially its Collaborative Activities, and assign-

ment of Community Process Objects to the process.

To graphically describe a Community Process,

an extension of Petri nets (Reisig, 1986) is used.

Petri nets are well known and are well suited for

modeling, analyzing and verifying process and data

flows. However, additional graphical elements are

required to describe human-centric activities espe-

cially communication behavior such as the behavior

in a Community Process. The graphical symbols of

the notation are listed in Table 2.

Table 2: Notation for Community Process modeling.

Symbol Description Meaning / Remarks

place Temporary storage for

Flowing Objects

transition Single Activity

directed arc Arcs running from a place to

a transition or vice versa

label Collaborative relationship;

A Collaborative Activity is

identified by a label on a

transition.

F-block Representation of an ab-

stract Finding Partner sub-

process

B-block Representation of an ab-

stract Building Relationship

sub-process

C-block Representation of an ab-

stract Collaboration Execu-

tion sub-process

block

arc

Block arc runing from F-

block to B-block or from B-

block to C-block

connection Assignment of a Communi-

ty User to a Single Activity;

connections may only exist

between unlabeled transi-

tions and members

member Representation of a Com-

munity User with a name

Besides these graphical extensions, a special re-

finement rule must be applied to labeled transitions

(Collaborative Activities) that is defined as follows.

In case of refinement of a labeled transition a se-

PROCESS-ORIENTED COORDINATION OF COLLABORATIONS IN SOCIAL NETWORKS

363

quence of F-block, B-block and C-block will be

created.

Figure 3 shows an example of a simple Commu-

nity Process that involves two network members

(with Name1 and Name2). Collaboration starts with

the Collaborative Activity CA.

Figure 3: Coordination of a research collaboration using a

Community Process.

For the modeling of Community Process Objects

UML class diagrams are used here. Figure 4 shows

the structure of the Community Process Object.

Figure 4: Structure of Community Process Object.

4 THE PROCESS OF MODELING

This section describes the process of modeling a

Community Process. The modeling process consists

of five steps.

1 First, Community Users have to be obtained

from e.g., the analysis of data from "event logs"

(van der Aalst and Song, 2004) or e-mails from

information and communication systems (Ya-

makami, 1998). Alternatively, user data from

existing social networks such as Facebook or

XING can beused. Based on the user’s relation-

ships the structure of the network can be

represented with a Sociogram given by Social

Network Analysis (Wasserman and Faust,

1994). Subsequently, statistical analysis me-

thods of the network structure can be used to de-

termine some important metrics for collabora-

tion such as centrality (a network member has a

lot of relationships to other network members),

indegree/outdegree (number of incom-

ing/outgoing connections in the role of reques-

tor and responder) and transitivity (two network

members who are both connected to a given

network member can be considered as directly

connected). These metrics are suitable to filter

contact persons or collaboration partners.

2 In the next step, all data objects describing

Community Content have to be specified such

as publication, collaboration agreement and ap-

pointment.

3 Generation of the Community Process based on

an activity list of a network member starting

from the first abstraction level. (Dengler et al.,

2009) suggested an approach how to automati-

cally generate a process model based on wiki

pages describing activities of an organizational

process.

4 Assignment of data objects from Community

Content to places and if necessary assignment of

Community Users to Single Activities.

5 Refinement of the Community Process using the

elements presented in Table 2 and modeling

more concretely the abstract Finding Partners,

Building Relationship and Collaboration Execu-

tion sub-processes.

The modeling steps 4-5, and possibly steps 1 and

2 are to be repeated until all Collaborative Activities

are defined and the Community Process is described

accurately enough.

5 USE CASE

In this Section, a use case of a Community Process

for the coordination of research collaboration will be

presented. The corresponding process on “collabora-

tion on writing a scientific paper” can be found in

(Klink et al., 2006). Figure 5 shows an example of

the network structure after the Community Process

Objects have been modeled. Note that this figure

shows only some of the edges with weights (com-

munication frequency of peers).

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

364

Figure 6: Creation of a Community Process model.

Figure 5: The network structure before the execution of

collaboration.

The next step is the generation of the first abstrac-

tion level of the Community Process. Subsequently,

the modeling steps 4-5, as explained in Section 4,

and possibly also steps 1 and 2 must be repeated. A

network member can either model the activities by

himself or use a modeling support tool

(Hornung et

al., 2008)

. The process will be modeled top-down.

Figure 6 shows the simplified Community Process.

In the execution period of the Community

Process, the communication details among the net-

work members (for example, A1, A6, A8, B1, C1

and C2), such as communication duration, frequency

and media will be collected and then analyzed as

mentioned in Section 4. Figure 7 shows an example

of the network structure at the end of the collabora-

tion. The structure includes some new relationships

(e.g., between B1 and C1 and C2) and arcs with in-

creased weights due to collaborative activities.

Figure 7: The network structure after the execution of

collaboration.

6 RELATED WORK

Related work can be found in three areas: (1) com-

bination of process modeling with social networks,

(2) Computer Supported Cooperative Work (CSCW)

and (3) eCollaboration approaches for supporting

scientific research.

In (Hornung et al., 2008) a recommendation-

based modeling support system for business

processes is described. To support the selection of

appropriate process models social networks are used

(Koschmider et al., 2009). (Khalaf et al., 2009) and

(Silva et al., 2009) are not using a recommendation

system based on social networks, but use social net-

works for an active exchange of process patterns.

The approaches above could be adapted in the mod-

eling of Community Processes in order to capture

target-specific process fragments.

PROCESS-ORIENTED COORDINATION OF COLLABORATIONS IN SOCIAL NETWORKS

365

Approaches in the areas of CSCW and eCollabo-

ration focus on the use of information and commu-

nication technologies to support collaboration. For

example in (Lubich, 1995), a CSCW framework for

scientific collaborations in Europe was described.

(Harrer et al., 2007) describes an approach in eCol-

laboration enabling researchers to detect interaction

patterns by utilizing logfiles of user actions captured

by system. (Luzón, 2009) considers the possibility of

using academic weblogs as tools for eCollaboration

to enable better communication among researchers.

The main innovation of the presented approach, in

contrast to the results from CSCW and eCollabora-

tion, is the flexibility and extensibility of the coordi-

nation of collaborations in social networks.

7 CONCLUSIONS AND

OUTLOOK

This paper proposes a process-oriented model for the

coordination of research collaborations in social

networks. Other available process-oriented ap-

proaches do not take into account the network (rela-

tionship) development appropriately, so that the

coordination of collaboration cannot take place effi-

ciently.

Coordination on the basis of the network devel-

opment in combination with Social Network Analy-

sis has the advantage that the activities and human

resources can be applied in an easier and more tar-

geted way for the initiation and execution of a colla-

boration. To organize collaboration the concept of

Community Process was introduced, which coordi-

nates both the individual and collaborative activities

of network members. The collaboration can be ef-

fectively controlled by a network member through a

Community Process because the communication

behavior with partners and the status of execution

are transparent.

Next steps of this work include the formalization

of all concepts of the Community Process in order to

obtain a system-supported execution of Community

Processes. An evaluation will be conducted investi-

gating the system's effectiveness and usability.

REFERENCES

Barnes, J. A., 1972. Social Networks, Addison-Wesley

Professional Press.

Benkler, Y., 2006. The Wealth of Networks: How Social

Production Transforms Markets and Freedom, Yale

University Press.

Dengler, F. et al., 2009. Collaborative Process Develop-

ment using Semantic MediaWiki. In WM’09: 5th

International Conference of Professional Knowledge

Management. Köllen Verlag.

Harrer, A., Hever, R., Ziebarth, S., 2007. Empowering

researchers to detect interaction patterns in e-

collaboration. In AIED’07 International Conference

on Artificial Intelligence in Education. IOS Press.

Hornung, T. et al., 2008. Recommendation Based Process

Modeling Support: Method and User Experience. In

ER’08, 27th International Conference on Conceptual

Modeling. Springer-Verlag.

Khalaf, R. et al., 2009. Enabling Community Participation

for Workflows through Extensibility and Sharing. In

BPMS2'09, 2th International Workshop on Business

Process Management and Social Software.

Klink, S. et al., 2006. A Service-oriented Information

System for Collaborativ Research and Doctoral Edu-

cation. In ICEBE'06, IEEE International Conference

on e-Business Engineering. IEEE Computer Society

Press.

Koschmider, A., Song, M., Reijers, H.A., 2009. Advanced

Social Features in a Recommendation System for

Process Modeling. In BIS’09, 12th International Con-

ference on Business Information Systems. Springer-

Verlag.

Lubich, H.P., 1995. Towards a CSCW Framework for

Scientific Cooperation in Europe, Springer-Verlag.

Luzón, M.J., 2009. Academic Weblogs as Tools for E-

Collaboration Among Researchers. In: E-

Collaboration: Concepts, Methodologies, Tools, and

Applications, Idea Group Publishing Press.

Reisig, W., 1986: Petrinetze: Eine Einführung, Springer-

Verlag. Berlin, 2

edition.

Schramm-Klein, H., 2005. Wettbewerb und Kooperation

in regionalen Branchenclustern. In Kooperationen,

Allianzen und Netzwerke: Grundlagen - Ansätze –

Perspektiven, Gabler-Verlag.

Silva, A.R. et al., 2009. AGILIPO: Embedding Social

Software Features into Business Process Tools. In

BPMS2’09 2th International Workshop on Business

Process Management and Social Software.

van der Aalst, W., Song, M., 2004. Mining social net-

works: uncovering interaction patterns in business

processes. In BPM’04, 2th International Conference

on Business Process Management. Springer-Verlag.

Wasserman, S., Faust, K., 1994. Social Network Analysis:

Methods and Applications, Cambridge University

Press.

Wenger, E., 1998. Communities of practice: learning,

meaning and identity, Cambridge University Press.

Yamakami, T., 1998. Social Process Visualization in Re-

gional Community Network Users. In APCHI 98, 3th

Asia Pacific Conference on Computer Human Interac-

tion. IEEE Computer Society Press.

WEBIST 2010 - 6th International Conference on Web Information Systems and Technologies

366