دیدگاه سرمایه اجتماعی روی مشارکت فرا دانش و محاسبات اجتماعی

Recent years have seen a substantial growth of social computing, where large numbers of individual users share content with others in online communities. Social computing systems have thus led to a profusion of highly heterogeneous data, further exacerbating the traditional problems of knowledge sharing. This has made Meta-knowledge (knowledge about knowledge) important and more widely used, as it helps users locate knowledge easily. However, the reasons for people's meta-knowledge contribution in the social computing context and the extent to which this may differ from traditional knowledge contribution remain largely unexplored. This gap is addressed in the present study. Building on social capital theory, and using a combination of survey and independent system data, we explore what affects individual meta-knowledge contribution on Flickr, a popular photo-sharing service.

Recent years have seen a substantial growth in social computing systems [22] and [36] that serve as intermediaries for social relations [46], and are characterized by online community formation and user content creation [12] and [36]. Some of the best known social computing systems are content sites such as Wikipedia, Flickr, YouTube, social networking sites such as MySpace and FaceBook, and social bookmarking services such as del.icio.us [28], [36] and [37]. Overall, these social computing systems are characterized by a high heterogeneity of information sources and make large amounts of information available to their users [45] and [54]: Wikipedia, for example, has more that 2 million entries in English alone [34], and Flickr has more than six billion uploaded photos. This accumulation of large amounts of heterogeneous information calls for the use of meta-knowledge – i.e. “the description of additional information about pieces of data stored in a database or knowledge base” [45] – in order to facilitate the organization and retrieval of this knowledge in the repository [30] and [45]. Thus, the rise of social computing in recent years has been supported by increasing use of user-contributed annotation in the form of tags [9], [14], [28] and [42], which convey knowledge about the content. Tags are keywords (e.g. “lighthouse”, “Christmas”, “California”) used to annotate various types of content, including images, bookmarks, blogs, and videos [42] and [49], and are a form of meta-knowledge that can significantly improve the discovery, retrieval, and understanding of relevant knowledge from the repository [45]. As tags provide knowledge about available content [14] and [29], organizers of content repositories often let users categorize content and share knowledge about it by tagging [9] and [29]. The popularity of tagging is attributed, at least in part, to the benefits users gain from effectively organizing and sharing very large amounts of information [9]. One prominent example of a tagging-intensive social computing system is Flickr, a photo sharing with over 51 million users [57]. Tagging in systems like Flickr is an important change in the way photos are organized and shared [49], as photos by themselves, despite their high information content, do not contain inherent information about what the photos represent, whereas the tags make the photos searchable by the uploading user, as well as by others [3]. While contribution of first-order knowledge has been studied, it is not clear how the results would apply to meta-knowledge contribution: when a person makes a first-order knowledge contribution by responding to a question posted on a bulletin board, or uploading a picture to a online photo-sharing community, the contribution is immediately visible and directly beneficial to other members; and the contributor has some level of certainty that her contribution will bring the reward she desired. Meta-knowledge, such as tagging, does not inherently bring any direct benefits to its audience. Although the trends and patterns emerging from a large collection of meta-knowledge in a community could be valuable information to certain parties, the immediate benefits expected by the contributor of individual pieces are achieved only when used in combination with first-order contribution. In addition, the effects of meta-knowledge contribution, while long-lasting, may not be as immediate as first-order contribution. While the benefits of first-order knowledge are clear to most members, how meta-knowledge contribution would benefit the contributors and others may not be clearly understandable to all members, especially novice members. For example, the content of a photo sharing web site such as Flickr benefits considerably from users adding meta-knowledge via tags especially as the first-order knowledge in this example is not easily amenable to searching. Therefore, from both theoretical and practical perspectives, it is important to investigate the antecedents to meta-knowledge contribution. In the present study, we aim to extend the use of Social Capital theory to explain contribution of Meta-knowledge, rather than first order knowledge. We draw on, and extend, recent work on the criticality of the structural, relational, and cognitive dimensions of social capital, as well as individual motivations as the starting point to investigate meta-knowledge contribution in online communities. Extending the qualitative work of [3], the present study contributes to the information systems literature on social capital by developing a framework in which the latter helps to understand a fast growing IS-facilitated phenomenon — social computing. Specifically, the findings make it possible to compare the relative strength of meta-knowledge contribution’ antecedents, and suggest that meta-knowledge contributors are primarily socially driven.

Given the growing interest among practitioners and researchers in meta-knowledge as means of facilitating the organization and sharing of large amounts of information [9], these communities may benefit from understanding what motivates users to contribute meta-knowledge. The present study is a first attempt, to the best of our knowledge, to address this question empirically and quantitatively in a social computing context. 6.1. Limitations Before we present the implications of our study, we need to note that there are several limitations of this study. First, our study applies a cross-sectional survey design; therefore, we cannot address the causality among the variables in our model. Although theoretical arguments can be made for the causal relationships between the independent variables (individual motivations and social dimensions) and the dependent variable (meta-knowledge contribution), our data analysis only permits us to demonstrate significant correlations in those hypothesized links. Future research can address this limitation by conducting a longitudinal study that tests these relationships. Second, our study was conducted on a specific social computing service – the Flickr online photo sharing community. Investigations of other types of social computing communities – bookmark tagging as in del.icio.us, video tagging as in youtube.com - can help verify the generalizability of our findings. In addition, future studies would help distinguish between different methods for tagging such as through the Flickr photo pages or through specific photo editing applications. Third, by focusing on users who uploaded photos to Flickr, users who uploaded more photos were more likely to be invited to take the survey. While the characteristics of respondents in this study were similar to characteristics of larger-scale samples of taggers in other studies [41], future research can help address this point by using other sampling methods. Future research can also be helpful in exploring the characteristics of non-taggers or users who only tagged once. Further, it should be noted that while actual system usage data come with advantages in alleviating common method bias, the measures we use by necessity, are limited by the usage data available and capture only one possible operationalization among many. For example, we measure meta-knowledge contribution by tracking the number of unique tags a user has assigned to all his/her photos. While this captures the quantity of contribution, other perceptual measures that capture quality of contribution could add further explanatory power to the research model. Similarly, cognitive capital is a complex construct that attempts to capture shared representations, interpretations and meanings among members of a social collective. In this study, we use respondent's tenure (number of years the user has been a member of Flickr) as a proxy for cognitive capital. While actual system data helps reduce measurement error, this trade-off also means that there is potential for future research to conceptualize and measure cognitive capital using richer operationalization. 6.2. Contributions This study extends previous theories of knowledge contribution by adopting it to study the factors that drive users to contribute meta-knowledge. To accomplish this, our research model uses recent work on the criticality of the three dimensions of social capital (structural, relational, and cognitive) and individual motivations as the starting point [32] and [55]. It then extends this social capital framework, which has been previously used to study knowledge contribution in organizational setting, to investigate meta-knowledge contribution in online social communities. Based on the qualitative work of [3], our research framework further develops appropriate set of individual level motivations that make sense in the meta-knowledge contribution context. A strength of the study is the use of data from multiple, independent sources. The data included survey responses as well as system log data retrieved directly through the Flickr API, to measure the dependent variable. This enabled us to overcome potential common method bias [50]. The study's findings have several implications for theory and practice: first, the study contributes to the body of information systems literature on social capital by developing a framework in which the latter helps to understand a new IS-facilitated phenomenon – social computing – and explain users’ perceptions and behaviors in this context. The concept of meta-knowledge has been studied in the form of computer generated metadata for large digital repositories, and its benefits in facilitating content discovery and retrieval is well accepted. Social computing services, on the other hand, enable individual users to be contributors and consumers of both content and meta-knowledge. Therefore, researchers also need to take a user-centric approach in understanding the dynamics of meta-knowledge generation and consumption in social computing environments. In this study we illustrated the role of social capital and individual motivations in user contribution of meta-knowledge. This understanding may serve as a useful first step for further research of social computing phenomena. Second, the study builds on insights identified in qualitative research [3], and extends them and enhances their generalizability, using quantitative methods, thus benefiting from both qualitative and quantitative perspectives. The quantitative approach in our study enables us to compare the relative strength of each antecedents of meta-knowledge contribution. We found the structural dimension of social capital, followed by motivation to benefit the public, to be the strongest predictors of contribution; while self motivation and cognitive dimension have least predicting power. Our results suggest that users of social computing systems are encouraged to contribute meta-knowledge mostly by the social aspects of these systems. This finding is particular interesting considering the fact that, compared to contributions of first order knowledge or content, contributions in meta-knowledge are much less visible to other members. In terms of implications for practice, the findings of the present study suggest that, as expected, both individual level motivations and the social capital dimensions affect users’ meta-knowledge contribution (as measured by users' tagging level). The findings therefore have implications for managers and organizers of social computing initiatives: it is advised that managers of collaborative content systems seeking to increase tagging activity focus their communication and marketing efforts on those factors that have a strong impact on the level of tagging. For example, the motivation of tagging photos for other users has a positive effect on tagging level. Therefore, organizers and managers of social computing systems are recommended to focus their marketing and user cultivation efforts in this area, by highlighting the possibility of being exposed to other, unknown users. Two of the three Social Capital dimensions were found to be significantly associated with tagging level. This highlights the importance of relational embeddedness and even more so, of structural embeddedness. Therefore, further opportunities for social interaction are essential and the inclusion of other avenues for social interaction as part of the design of such systems need to be encouraged. Further, beyond the systems’ design, social interaction can be increased by encouraging users to use existing social interaction avenues. While the research model developed in the present study investigates social tagging systems in personal, non-work setting, the model can be adapted to study what motivates people to contribute meta-knowledge in organizational settings, by developing an appropriate theory for individual motivations and organizational settings in that context. Further research may also be helpful in understanding how different motivations influence contribution in different content sharing systems. In addition, future research can help to explore the interrelationships among the three social capital dimensions, which was regarded as an important area for future research by [32]. Another avenue for further exploration is the examination of the strength of social ties [10] and [11] in social computing initiatives, and its role in explaining meta-knowledge contribution. Social computing is a fast growing phenomenon, which has been, to date, largely under-explored. The present study, addressing one of the prominent social computing systems, is hopefully a useful step in this direction.