Computers & Education 54 (2010) 12331240

Contents lists available at ScienceDirect

Computers & Education

journal homepage: www.elsevier.com/locate/compedu

Promoting social network awareness: A social network monitoring system

Rita Cadima a,*, Carlos Ferreira a, Josep Monguet b, Jordi Ojeda b, Joaquin Fernandez b
a b

Polytechnic Institute of Leiria, Rua General Norton de Matos, Apartado 4133, 2411-901 Leiria, Portugal Polytechnic University of Catalonia, Av. Diagonal, 647, 08028 Barcelona, Spain

a r t i c l e

i n f o

a b s t r a c t
To increase communication and collaboration opportunities, members of a community must be aware of the social networks that exist within that community. This paper describes a social network monitoring system the KIWI system that enables users to register their interactions and visualize their social networks. The system was implemented in a distributed research community and the results have shown that KIWI facilitates collecting information about social interactions. Furthermore, the visualization of the social networks, given as feedback, appeared to have a positive impact on the group, augmenting their social network awareness. 2009 Elsevier Ltd. All rights reserved.

Article history: Received 12 March 2009 Received in revised form 9 November 2009 Accepted 17 November 2009

Keywords: Awareness Social networks Distributed learning environments Multimedia/hypermedia systems

1. Introduction Knowledge is created and exchanged to a large extent through informal social interactions (Ogata, Yano, Furugori, & Jin, 2001; StorbergWalker & Gubbins, 2007) that allow the transfer of sensory information, intuition and non-verbal communications (Cummings & Yeng, 2003). Knowledge ows depend on the connections between individuals and on their attitude about sharing knowledge (Inkpen & Tsang, 2005; Ipe, 2003; Lin, 2007; Wang & Yang, 2007). Supporting collaboration and work in these social networks has been increasingly recognized as important for organizations to compete on knowledge and on their ability to innovate and adapt (Cross, Parker, & Borgatti, 2002), calling attention to the importance of considering the social capital. Social capital refers to the collection of social trust, norms and networks that people can draw upon to solve common problems. While human capital refers to properties of individuals such as knowledge, social capital implies connections among individuals and the value derived from these connections (Daniel, McCalla, & Schwier, 2002). In distributed communities, communication technologies alone seem not enough to promote communication and knowledge sharing (Cummings & Yeng, 2003; Lin, 2007). It appears to be very important to be aware of others in order to communicate and collaborate (Hu, Kuhlenkamp, & Reinema, 2002). Thus, virtual environments must provide means to communicate social cues and context information (Kreijns, Kirschner, Jochems, & van Buuren, 2007). Supporting awareness to be aware of the ideas, knowledge, and activities of the others has been used as one of the strategies to increase knowledge sharing and collaboration opportunities (DiMicco, Hollenbach, Pandolfo, & Bender, 2007; Gutwin & Greenberg, 1997; Ogata & Yano, 1998). Awareness systems help people to effortlessly maintain awareness of others, thus facilitating lightweight, emotional, and informal forms of communication (Van Baren, IJsselsteijn, Markopoulos, Romero, & de Ruyter, 2004). Different mechanisms were applied to build awareness of who knows what by distributing information about peoples expertise and it has proven effective in increasing knowledge awareness (Cross, Parker, Prusak, & Borgatti, 2001). But knowledge in communities is highly implicit and socially constructed (Novak & Wurst, 2005) and knowing that someone else knows something of relevance does little good if people cannot gain access to their knowledge and help just in time. This accessibility is directly connected to social network awareness, which we understand as the awareness of social relationships within the group the awareness of who knows whom. It seems helpful to map access relations at a network level to understand who is able to reach whom in a sufciently timely way (Cross et al., 2001). In a virtual environment users must be able to perceive and compare the social patterns of activity to their own models of work and interaction. This could enhancing users motivation to communicate and collaborate and will allow them to structure their social networks to maximise their benets by getting closer to the existing resources and opportunities (Soller, Martnez, Jermann, & Muehlenbrock, 2005).
* Corresponding author. Address: Polytechnic Institute of Leiria, Campus 1, Rua Dr. Joo Soares, Apartado 4045, 2411-901 Leiria, Portugal. Tel.: 00351 938 256607/00351 244 829400; fax: 00351 244 829499. E-mail addresses: rcadima@esel.ipleiria.pt (R. Cadima), cjrf@estg.ipleiria.pt (C. Ferreira), jm.monguet@upc.edu (J. Monguet), jordi.ojeda@upc.edu (J. Ojeda), jfernandez@upc.edu (J. Fernandez). 0360-1315/$ - see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.compedu.2009.11.009

1234

R. Cadima et al. / Computers & Education 54 (2010) 12331240

Fig. 1. Monitoring system model.

However, only a subset of all the interaction occurring in the workspace are usually monitored. Considering the importance of awareness, it seems relevant to explore which techniques can be used to support it (Otjacques, Noirhomme, Gobert, & Feltz, 2006). Also, according to Zheng and Yano (2007), more efforts are required to develop tools to uncover social networks and explore this social dimension of awareness. The present study addresses these issues by describing a social network monitoring system expected to enhance social network awareness in a distributed community. We are particularly interested in knowledge intensive domains, given that scientic work processes can be seen as informal learning processes with a high level of social interaction that allows knowledge sharing and knowledge construction (Braun, Schmidt, & Hefke, 2007). In a scientic research domain communication and sharing is essential and could facilitate knowledge divulgation and expert accessibility. For example, a researcher more advanced in a specic eld could give some useful cues to a beginner student. Monitoring knowledge sharing could increase the comprehension about how much each individual receives from the community and how he or she is using knowledge from within the community. Taking this into account, we developed a social network monitoring system aimed at uncovering the social network of a distributed R&D community knowledge interactions to work and innovate (KIWI)1 that addresses simultaneously: (a) gathering information about social networks and (b) promoting social network awareness. This system innovates by asking directly people about their interactions, allowing them to register every kind of interactions. The purpose of this paper is to present and describe the implementation of KIWI system in a real world environment. It also intends to analyse its effects on users social awareness and behaviour. More specically, in order to address these goals, preliminary data collected during its implementation as well as the evaluation of the system by the users are analysed and further discussed.

2. Overview of KIWI system KIWI system is a social network monitoring system that depends on active participation of users in the data gathering process. According to system model (see Fig. 1), KIWI is a web-based application with two separate views: one for data collection and other for feedback. The system provides users with a gathering tool for registering their interactions and automatically analyses and presents social network information through a visualization tool. Explicit social network information is extracted from a database through social network analysis (SNA) techniques. SNA provide a rich and systematic means of assessing informal networks by mapping and analyzing relationships among people (Cross et al., 2001) and can be a valuable analytical tool to examine complex social processes and outcomes in communities (Cho, Gay, Davidson, & Ingraffea, 2007). In addition to its potential to go further in a systematic analysis of social network by researchers and/or community managers, the system supports social network awareness of users by making the hidden networks visible to all community, without abstracting or evaluating users behaviours. In this way users could be more aware of the connections that allow knowledge transfer and the possibilities to access knowledge. By directly asking users about their interactions it is possible to monitor every kind of interaction, from face-to-face meetings to mail and chat interaction, without implying major changes to users current behaviour (the imposition of new communication tools could change the existing spontaneous informal network and would not ensure that all of what was happening was being recorded). Although the required involvement in the data gathering process creates additional workload for users, potentially leading to a disparity between effort and benet (Rittenbruch, Viller, & Manseld, 2007), we note two advantages of this strategy. First, this option can act as a ltering strategy which will increase the extraction of meaningful information and decrease the burden in analysis, instead of producing extensive data as most monitoring systems do, which in turn would require considerable effort to uncover signicant relationships within the group (Chen, Wang, & Ou, 2003). Second, this strategy is likely to promote individual responsibility, to strengthen trust among participants, and to improve self-awareness, self-direction and self-management of their own activities (Zheng, Li, Ogata, & Yano, 2007).
1 It should be mentioned that there is a different project with the same name KIWI. This EU Project is concerned with knowledge management in semantic wikis and it was funded by the European Commission under the Project Number 211932.

R. Cadima et al. / Computers & Education 54 (2010) 12331240

1235

2.1. Data gathering tool The main requirement for the design of this data gathering tool was to minimize the additional workload for users. It is implemented through a simple form-based web page where a user sees a list of community members (identied by name and picture) and responds by clicking on those people with whom he/she has interacted (see Fig. 2). At the top of page there are the questions and denitions that explain the interactions that should be registered. For each person, the user can identify different kinds of interactions. The layout of the interface adapts with ongoing use. After each users rst response, the community members selected in the previous sessions appear in a prominent area my network. In this way the effort for looking for regular co-workers is reduced and, at the same time, the user can be more aware of his/her regular network. To implement this tool, two things are necessary: (a) identify the community members that will be monitored and (b) dene what kind of interactions will be registered. After establishing the community participants it would only be possible to register interactions between these people. This data gathering tool was already tested in a previous study (Cadima et al., 2008) allowing us to assess its usability and giving an opportunity to collect users opinions about which social network information was most relevant to them. These rst results guided the design of the visualization tool. 2.2. Visualization tool The visualization tool is a web page that contains weekly updated diagrams and real-time bar charts. Social network diagrams are used to visually represent networks and uncover patterns of peoples interactions (see Fig. 3 in Section 3.2). To let users assess the effectiveness of their personal network, the visualization tool also provide graphical quantitative information trough bar charts (number of people in their individual network, frequency of each type of interactions). The visualization tool works automatically from the system database, and there is great exibility when deciding what kind of information will be displayed. More details about the options taken for this tool in the eld test will be provided in Section 3.2.

Fig. 2. Screenshot of the data gathering tool.

1236

R. Cadima et al. / Computers & Education 54 (2010) 12331240

Fig. 3. Screenshot of the visualization tool.

3. Method 3.1. Participants This study was developed within a distributed community of 37 researchers the Multimedia Engineering PhD Programme of Polytechnic University of Catalonia (UPC). In this community, many Research, Development and Innovation (RD & I) projects and services rely upon

R. Cadima et al. / Computers & Education 54 (2010) 12331240

1237

multidisciplinary teams that bring together different expert knowledge domains (engineers, designers, teachers, mathematicians, anthropologists, psychologists). There is a central unit located in Barcelona, Spain, but many members are located on other countries (Venezuela, Mexico, Colombia, Portugal, Denmark, and USA) and primarily maintain virtual interaction with others. This community uses a web platform for information sharing and there are weekly seminars (virtual conferences) for individual and group research presentations. Most communications occur outside this platform through mail, chat, or, in some cases, face-to-face encounters. 3.2. Procedure A preliminary study in this community allowed understanding the work methods and typical research activities of each individual, as well as to describe existing networks of communication and collaboration in terms of tools used, patterns, frequency, type of information shared and existing help. Community members were also asked about their satisfaction with the communication and collaboration occurring and it was observed that the participants longed for more interaction. These results, by showing to communitys managers the need for strategies to improve communication and collaboration opportunities, gave support to the implementation of KIWI system in this community. To implement KIWI system three steps were necessary: (a) identify which people should be monitored; (b) dene what kind of interactions should be registered; (c) dene what kind of information should be ltered out from the system database and showed back to the community members. After the rst diagnostic and several meetings with supervisors and community directors, it was decided that all the active members of community (37 people) should be monitored and that they will be weekly asked about the interactions that allowed knowledge transfer. The system was integrated into the community web platform COLS (www.e-cols.net) and the participants were asked to respond to the KIWI data gathering tool every week, identifying those people with whom they interacted for knowledge sharing during that week. Each user had to classify the giving or receiving (or both) character of the interaction with each other user. During the 8 weeks of the eld test the visualization tool provided two types of information (see Fig. 3): (a) network diagrams and (b) graphical quantitative information. (a) Network diagrams: We used social network analysis software tool NetDraw (Borgatti, 2002) to visually represent all the social connections registered. Every week a new diagram was posted, presenting the social network accumulated in the past weeks. In these network diagrams, each node in the network represents a person and each arc represents a knowledge transfer channel (see Fig. 3). Arrows were used to reect the direction of knowledge transfer. Every knowledge transfer was represented, even if it was only registered by the sender or by the receiver. Colours and symbols were used on nodes to give meaningful information: triangles represented supervisors while PhD students were represented with circles and squares, according to their status as researchers; the Barcelona group was in blue and everyone else in red. (b) Quantitative information: Six graphs were presented with weekly and cumulative information: number of contacts, number of knowledge receiving transfers and knowledge giving transfers. Because the visualization tool is personalized and adapts to the user, both individual values and group values were presented to each user. Group values were presented to distinguish local group social activity (people located at Barcelona) from distance group activity (people located at disperse countries) and to distinguish beginnerstudents from advanced students and supervisors. After using KIWI for eight weeks, users were requested to ll out an on-line survey to evaluate system usability and the effects of its usage on users social network awareness and perceived motivation to interact with others. The questionnaire had 13 items (e.g., there was a few time effort to contest the KIWI every week) with a 5-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = agree/disagree, 4 = agree, 5 = strongly agree), and the participant was asked to rate each item. Items 14 covered the efciency of the system (e.g., I could understand all the information displayed in the KIWI system), items 58 were used to evaluate the effect of the system on users social network awareness (e.g., the information displayed improved my awareness about the others and their interactions), and items 911 focused on users perceptions about the effects of the system on their motivation to communicate and collaborate (e.g., my participation in this study gave me more motivation to help others).

4. Results 4.1. KIWI system usage We use system logs to describe KIWI system usage. During the eld test, the gathering tool was used a mean average of 4.5 times per person during the 8 weeks. The individual mean average was 6.8 interactions by person by week (SD = 4.4), and the time average of each response was 1.86 min. These results indicate how easy was to manipulate the gathering tool. Besides registering their interactions once a week, users were invited to access the visualization tool for receiving feedback on communitys interactions. The visualization tool was used 9.92 times per person during the 8 weeks, showing that users often accessed KIWI just to visualize their social network. 4.2. Social network analysis In this section we present a small amount of the total data collected to illustrate the potential of the system on given relevant and useful information about the communitys social network. The total data registered by participants allowed the description of the connections and members centrality within the network. We use SNA tools UCINet (Borgatti, Everett, & Freeman, 2002) and NetDraw (Borgatti, 2002) for representing and analyzing the collected data. Diagrams (A) and (B) in Fig. 4 display all connections registered, showing a cohesive network

1238

R. Cadima et al. / Computers & Education 54 (2010) 12331240

Fig. 4. Social networks for knowledge sharing.

(density = 0.2442) with no isolated sub-groups except a single individual (he answered KIWI gathering tool several times informing that there were no interactions). To better understand the structural importance and prominence of each person, nodes are sized according to individuals degree (diagram (A)) and betweenness (diagram (B)). Degree refers to the extent to which an individual has numerous connections to other members. High degree centrality seems important, because it has been shown that it is positively associated with performance through the improvement of individuals access to resources (Cho et al., 2007). Betweenness captures the property of frequently lying along the shortest path between pairs of persons. In diagram (B) large nodes identify those people who constitute access bridges for those who are not directly connected. Members occupying these positions seem to have more control over diverse resources located in multiple sub-groups (Cho et al., 2007). Our results show that several people from the local group (blue)3 act as a bridge, but there are also four people at distance group (red) with a central role in networks accessibility. Diagram (C) in Fig. 4 displays only distance group network. Inside this subgroup, network cohesion decreases (density = 0.149) and an isolate two persons group shows up. Diagram (D) displays local group network revealing an extremely cohesive (density = 0.552) and non-hierarchical structure. The central position of local group in community social network may represent to this group a greater advantage in receiving and applying knowledge (Soreson, Rivkin, & Fleming, 2006).

4.3. Accuracy of users perceptions on knowledge sharing In the previous section we described the connections between people (who interact with whom). In this section we analyse more deeply the frequency and direction of knowledge sharing, because in each two-people connection, there could be several knowledge transfers matching to each week of the study. There were a total of 633 knowledge transfers resulting from 549 records of knowledge receiving and 339 of knowledge giving. These results show that users tend to be more willing to acknowledge that they had received. We analyse how many transfers were registered by both giver and receiver and this accuracy of users answers is displayed in Table 1. Within all community, there were only 40.3% transfers registered by both giver and receiver. These results are signicantly different from the results obtained in the rst pilot test where 60.5% of knowledge transfers were identied by both giver and receiver within a group of 15 people for a 4 weeks period trial. Exploring if these values could be explain by the participation rate, we analysed the transfers occurred only within a subgroup of 16 participants with a higher rate of participation (responded six or more weeks to KIWI) and the accuracy of answers gets slightly higher, with 55.9% registered by both giver and receiver. We also explored if face-to-face interactions could be a confusing factor when deciding if there was a knowledge transfer occurring. However, considering only the interactions where atleast one of the members belong to the distance group (i.e., ignoring transfers between two local group members), the results were very similarly to those found in all community, with 41.1% registered by both giver and receiver. The potential ambiguity existing in the gathering process could be explained by the nature of the question asked: With whom did you interact for knowledge sharing? The act of identifying when there exists knowledge sharing could de complex and ambiguous, especially when sharing tacit knowledge (Chiva & Alegre, 2005; Garcia-Perez & Mitra, 2007). In these cases it is easy for the receiver acknowledge that knowledge transfer occurred but difcult for the giver to realize that there was a useful knowledge transfer. Table 2 displays the mean ranks by groups and the proportion of interactions for knowledge receiving and knowledge giving. The comparison of groups results shows that people from local group had a higher level of interaction, in both receiving and giving. Supervisors seemed to have higher levels of interaction and were more likely to assume knowledge giving. Beginner-students had the lowest rates, especially on giving knowledge. The high values of the standard deviations suggest that there was a great variability of behaviours among each group.
2 3

Network density is the proportion of lines present in the graph to the maximum number of lines possible and it is often interpreted as a measure of cohesion of the group. For interpretation of color in Figs. 24, the reader is referred to the web version of this article.

R. Cadima et al. / Computers & Education 54 (2010) 12331240 Table 1 Percentage of knowledge transfers registered in each of three situations: registered by both receiver and giver, registered only by receiver or registered only by giver. Knowledge transfers between all members (n = 633) Registered by both giver and receiver Registered only by receiver Registered only by giver 40.3 46.4 13.3 Knowledge transfers between members with higher participation rate (n = 186) 55.9 31.7 12.4

1239

Virtual knowledge transfers (n = 438) 41.1 46.1 12.8

Table 2 Mean ranks by groups and proportion inside group of interactions for knowledge receiving and knowledge giving registered by each individual of the group. Group n Knowledge receiving Mean Local group Distance group PhD beginner-students PhD student-researchers Supervisors N 14 23 10 22 5 37 16.71 13.69 11.80 15.27 19.0 14.83 SD 9.8 11.8 8.69 11.3 8.44 11.0 % 58.8 64.3 73.3 60.5 55.2 61.8 Knowledge giving Mean 11.71 7.61 4.30 9.95 15.4 9.16 SD 8.0 9.2 5.7 9.3 8.4 8.9 % 41.2 35.7 26.7 39.5 44.8 38.2

Table 3 Survey results on evaluating the systems usability and the effects of its usage in users social network awareness. N 1. There was a few time effort to contest the KIWI every week 2. I had to make hardly any intellectual effort to contest KIWI 3. I had to make hardly any intellectual effort to understand the information about the social networks 4. I could understand all the information displayed in the KIWI system 5. The reection I had to make to contest KIWI made me more aware about my interactions and my role in the community 6. The information displayed was relevant for me 7. The information displayed improved my awareness about the others and their interactions 8. The information displayed improved my awareness about my interactions and my role in the community 9. My participation in this study gave me more motivation to interact with others 10. My participation in this study gave me more motivation to help others 11. My participation in this study gave me more motivation to ask for help 12. I found positive and useful the reection I had to make to contest KIWI 13. Im satised with my participation in this study 23 23 22 22 23 22 23 23 23 23 22 23 23 Negative (%) SD + D 4.3 0 18.2 27.3 4.3 4.5 4.3 8.7 17.4 17.4 18.2 4.3 0 Positive (%) A + SA 95.6 95.6 63.4 59.1 87 68.2 69.5 73.9 56.5 56.7 45.4 73.9 73.9 Strongly agree 65.2 65.2 4.5 9.1 52.2 18.2 21.7 39.1 8.7 4.3 13.6 39.1 39.1 Mean 4.57 4.61 3.45 3.36 4.35 3.77 3.87 4.04 3.48 3.43 3.41 4.09 4.13 SD 0.73 0.58 0.96 1.09 0.83 0.92 0.81 0.98 0.90 0.84 0.86 0.90 0.81

N number of answers; SD + D percentage of strongly disagree and disagree; A + SA percentage of agree + strongly agree; % SA percentage of strongly agree; M mean average of response; SD standard deviation.

4.4. Users social network awareness After the eld test, users were request to answer to an on-line survey that intended to evaluate the usability of the system and the effects of its usage in users social network awareness and behaviours. Table 3 summarizes surveys results. Regarding the evaluation of the system efciency (items 14), about 95% of users answered that the time and intellectual effort required to contest KIWI was hardly any (items 1 and 2), conrming that the data gathering tool was very simple and easy to use. However, some users noted some effort and difculty in understanding all the information displayed in the visualization tool (items 3 and 4), which indicates that the graphics provided could be slightly complex or that too much information was being given. This difculty could have some negative effects in the reection and interpretation process and deserve future attention and improvement. In respect to the effectiveness of the system in augmenting users social network awareness (items 58), the participants acknowledged an improvement on their awareness. This seems especially signicant when users were registering their interactions (item 5). We also analysed the users perceptions about systems effects in their motivation to communicate and collaborate (items 911). According to participants opinion, the system had a slight impact in users motivation. About 57% of users considered that their motivation to interact and help others had improved. A little less, 45%, admitted to have more motivation for asking help. In a general way, users were very satised in using KIWI system and considered that the reection required was positive and useful and the information provided was relevant to them. 5. Conclusions This paper has presented KIWI as a social network monitoring system and its application to one real world scenario. Results have shown that users can easily use KIWI to give information about their social networks of knowledge sharing and that the collected data allowed

1240

R. Cadima et al. / Computers & Education 54 (2010) 12331240

displaying relevant information about these networks. Users considered positive and useful the reection they had to make when using KIWI and acknowledged an improvement on their social network awareness. The analysis of the collected information revealed some interesting patterns inside the community, as the non-hierarchical structure of Barcelona group and its central position in the community network; or the identication of a few members acting as access bridges between people that are not directly connected. It was also observed that users tended to be more aware when receiving from others than when giving. These results deserve further attention in the future and more deepen analysis seems necessary to understand communitys interactions patterns. Although with only 40% of accuracy in responses (transfers registered by both giver and receiver), the option of using participants perceptions appears to have the advantage that users select and register the interactions they believe to be signicant, producing a meaningful and self-relevant explicit social network. This study had some limitations that should be acknowledged, namely the reduced time of implementation and the participation of only one community. KIWI could be implemented in organizations that, while not based on teams work and group tasks, desire strategies to improve knowledge sharing. Further work based on other real world scenarios is needed to validate the system versatility in adapt to diverse communities. Also there was not sufciently control on measuring how users interpret the information received and there were reports on difculty in understanding all the social network information. This fact points out the need of future work to improve the effectiveness of the visualization tool. And, once social structures change over time, as do their effects on individuals (Cho et al., 2007), further and longer research is also needed to explore the effects of social network awareness in promoting communication and collaboration. References
Borgatti, S. P. (2002). Netdraw network visualization. Harvard, MA: Analytic Technologies. Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for windows: Software for social network analysis. Harvard, MA: Analytic Technologies. Braun, S., Schmidt, A., & Hefke, M. (2007). A socially-aware desktop for e-science: Supporting learning in networked scientic processes. In Norbert Gronau (Ed.), 4th conference on professional knowledge management experiences and visions (WM 2007), workshop on collaborative knowledge management (CoKM) (pp. 4754). GITO. Cadima, R., Ferreira, C., Ferruzca, M., Monguet, J., Ojeda, J., & Fernandez, J. (2008). Knowledge sharing in a distributed R&D community: an experience around social awareness. In G. Richards (Ed.), Proceedings of world conference on e-learning in corporate, government, healthcare, and higher education (pp. 644649). Chesapeake, VA: AACE. Chen, G. D., Wang, C. Y., & Ou, K. L. (2003). Using group communication to monitor web-based group learning. Journal of Computer Assisted Learning, 19, 401415. Chiva, R., & Alegre, J. (2005). Organizational learning and organizational knowledge: Towards the integration of two approaches. Management Learning, 36(1), 4968. Cho, H., Gay, G., Davidson, B., & Ingraffea, A. (2007). Social networks, communication styles, and learning performance in a CSCL community. Computers & Education, 49, 309329. Cross, R., Parker, A., & Borgatti, S. P. (2002). Making invisible work visible: Using social network analysis to support strategic collaboration. California Management Review, 44(2), 2546. Cross, R., Parker, A., Prusak, L., & Borgatti, S. P. (2001). Knowing what we know: Supporting knowledge creation and sharing in social networks. Organizational Dynamics, 30, 2, 100120. Cummings, J., & Yeng, B. S. (2003). Transferring R&D knowledge: The key factors affecting knowledge transfer success. Journal of Engineering and Technology Management JETM, 20, 3968. Daniel, B., McCalla, G., & Schwier, R.A. (2002). A process model for building social capital in virtual learning communities. In Proceedings of the international conference on computers in education (pp. 574577). Auckland, NZ. DiMicco, J., Hollenbach, K., Pandolfo, A., & Bender, W. (2007). The impact of increased awareness while face-to-face. HumanComputer Interaction, 22, 4796. Garcia-Perez, A., & Mitra, A. (2007). Tacit knowledge elicitation and measurement in research organisations: A methodological approach. The Electronic Journal of Knowledge Management, 5(4), 373386. Gutwin, C., & Greenberg, S (1997). Workspace awareness. Position paper for the ACM CHI97 workshop on awareness in collaborative systems (pp. 2227), McDaniel & Brinck, Atlanta, Georgia, March 1997. Hu, B., Kuhlenkamp, A., & Reinema, R. (2002). Supporting group awareness in web-based learning environments. In Y. Han, S. Tai, & D. Wikarski, (Eds.), EDCIS 2002, Lecture notes in computer science (Vol. 2480, pp. 525536). Inkpen, A., & Tsang, E. (2005). Social capital, networks, and knowledge transfer. Academy of Management Review, 30(1), 146165. Ipe, M. (2003). Knowledge sharing in organizations: A conceptual framework. Human Resource Development Review, 2(4), 337359. Kreijns, K., Kirschner, P., Jochems, W., & van Buuren, H. (2007). Measuring perceived sociability of computer-supported collaborative learning environments. Computers & Education, 49, 176192. Lin, H. F. (2007). Knowledge sharing and rm innovation capability: An empirical study. International Journal of Manpower, 28(3/4), 315332. Novak, J., & Wurst, M. (2005). Collaborative knowledge visualisation for cross-community learning. In S. O. Tergan & T. Keller (Eds.). Lecture notes in computer science (Vol. 3426, pp. 95116). Springer. Ogata, H., & Yano, Y. (1998). Knowledge awareness: Bridging learners in a collaborative learning environment. AACE, International Journal of Educational Telecommunications, 4(2/3), 219236. Ogata, H., Yano, Y., Furugori, N., & Jin, Q. (2001). Computer supported social networking for augmenting cooperation. Computer Supported Cooperative Work, 10(2), 189209. Otjacques, B., Noirhomme, M., Gobert, X., & Feltz, F. (2006). Cooperation indexes to support workspace awareness. In 12th international workshop, CRIWG 2006. Lecture notes in computer science (Vol. 4154, pp. 94101). Rittenbruch, M., Viller, S., & Manseld, T. (2007). Announcing activity: Design and evaluation of an intentionally enriched awareness service. HumanComputer Interaction, 22, 137171. Soller, A., Martnez, A., Jermann, P., & Muehlenbrock, M. (2005). From mirroring to guiding: A review of state of the art technology for supporting collaborative learning. International Journal of Articial Intelligence in Education, 15, 261290. Soreson, O., Rivkin, J., & Fleming, L. (2006). Complexity, networks and knowledge ow. Research Policy, 35, 9941017. Storberg-Walker, J., & Gubbins, C. (2007). Social networks as a conceptual and empirical tool to understand and Do HRD. Advances in Developing Human Resources, 9(3), 291311. Van Baren, J., IJsselsteijn, W., Markopoulos, P., Romero, N., & de Ruyter, B. (2004). Measuring affective benets and costs of awareness systems supporting intimate social networks. In A. Nijholt & T. Nishida (Eds.), Proceedings of the social intelligence design. CTIT workshop proceedings series WP04-02 (pp. 1319). Wang, C. C., & Yang, Y. J. (2007). Personality and intention to share knowledge: An empirical study of scientists in an R&D laboratory. Social Behavior and Personality, 35(10), 14271436. Zheng, Y., Li, L., Ogata, H., & Yano, Y. (2007). Support online social interaction with context-awareness. International Journal of Continuous Engineering Education and Lifelong Learning, 17(2/3), 160177. Zheng, Y., & Yano, Y. (2007). A framework of context-awareness support for peer recommendation in the e-learning context. British Journal of Educational Technology, 38(2), 197210.