Collaborative work across boundaries: Filtering, tagging and rating shared resources in interorganizational settings Polychronis Papadakis and Demosthenes Akoumianakis Department of Informatics Engineering Technological Education Institution of Crete (TEI-C) Heraklion, Crete, Greece
[email protected] ;
[email protected] Abstract — The present work concentrates on collaborative virtual work and the tactics for spanning organizational boundaries. To this effect, it proposes an architectural pattern for Collective Collaborative Tagging (CCT) that integrates cloud-based resource sharing and enterprise portals. The current implementation exploits Dropbox for file sharing and the Liferay enterprise portal for managing meta-data and collective wisdom. The results illustrate the feasibility of a tactic for spanning inter-organizational boundaries to establish and sustain improved information connectivity between members of cross-functional virtual teams. They also point to capabilities of digital trace data to untangle intrinsic properties of enacted cyber-structures resulting from the users’ recurrent co-engagement. Keywords—inter-organizational collaboration, resource sharing, tagging, enacted cyber-structures
I.
INTRODUCTION
Computer-mediated collaborative work has been studied extensively in many different scholarships and from a variety of perspectives. Our interest in this paper focuses on eCollaboration [1] and the technologies used to span organizational boundaries in order to facilitate cross- or interorganizational engagements [2][3]. It turns out that crossorganizational collaboration sets targets that are far beyond conventional collaborative functions such as distributed file management, conferencing, shared repository management and document exchanges. Instead, it fosters streamlining of autonomous and situated organizational activities with a boundary core that makes them intelligible in contexts of distributed organizing. Nonetheless, such a target represents a non-trivial undertaking. Breakdowns emerge not only due to geographic or temporal differences, but also from variation in terminology, symbol sets and/or organizational structure and norms [4]. Reaching out to attain shared goals requires that information presented to actors is not only structured and meaningful, but also intelligible so as to be applicable to local decision making and the context of each participating organization. This makes compelling the need for systems that
are designed to exhibit the capacity for organizational boundary spanning through tactics that foster computational off-loading, activity awareness and improved capacity for processing knowledge-based boundary objects and artifacts. Traditionally, research into inter-organizational systems (IOS) has underestimated the complexity of this undertaking, emphasizing primarily static demarcations of boundaries and transactionoriented operations. As a result, the relevant literature offers very limited insight into how shared ground, social ties and togetherness may be fostered through joint workflows that cross borders and span institutional settings [5]. The present work is a step in the direction of promoting interorganizational collaboration using web 2.0 concepts and in particular certain technologies such as cloud-based file sharing, social tagging systems and enterprise bookmarking that appear to be very promising. Through this lens, we aim to approach some of the challenges involved in establishing distributed and interorganizational systems that capitalize on social interactions of ambient affiliates and shared resources. Specifically, our focus is on establishing and sustaining an „enacted‟ form of information connectivity across boundaries by exploiting appropriate boundary negotiating tactics, thus contributing to a more detailed understanding of how human and material agencies are entangled to constitute increasingly synthetic worlds. The rest of the paper is structured as follows. The next section presents the theoretical streams that guide the present effort. This, amongst other things, anchors related works and motivates the research focus of the present research. Then, we describe the architectural pattern of a system that exploits social tagging, enterprise bookmarking and digital trace data in inter-organizational settings. The paper is concluded with a discussion on how the system is deployed in a research setting and reflections on future work. II.
BACKGROUND AND RELATED WORK
The present research links with scholarships on IOS and boundary spanning, especially the tactics offered by web 2.0 and social computing technologies.
A. Inter-organizational systems Kumar and van Dissel [6] introduced the term Interorganization Information Systems (IOS) to coin a particular genre of “… technologies designed and implemented to operationalize the relationships between partners in an alliance” (p. 8). Today the term is used widely to qualify a range of technologies that allow cooperation between businesses or institutions to attain a variety of shared objectives [7]. Nonetheless and in spite of rising research attention, extant interorganizational exchange research involving ties or networks is relatively limited [8]. Consequently, there are important gaps in understanding the implications of IOS on connectivity and enacted cyber-formations. To this effect, two broad research areas are still pending attention. The first relates to the formation and management of inter-organizational ties in digital space, while the second amounts to the type and role of boundaries and the corresponding boundary spanning tactics deemed as appropriate. Regarding IOS and ties, scholarship on IOS implementation (both in business and public sectors) tends to concentrate at the levels of e-commerce and/or e-operations where ties between agents are typically strong and rigid. In most cases, cooperation is grounded on the basis of a shared suite of tools (i.e., airline ticketing), an industry standard (i.e., EDI) or a common practice (i.e., journalism). The more challenging issue of e-Collaboration between agents with weak ties favoring local autonomy, continuous learning and exploratory engagements is less frequently addressed. Furthermore, the dynamics of ties (e.g., weak ties evolving to strong ties and vice versa) is another area pending further research and empirical insight. With respect to boundaries, it is common for IOS to take as granted certain types and forms of boundaries that have a somewhat static demarcation (e.g., temporal or spatial distance of peers). Although the relevant literature acknowledges the relevance of dynamic knowledge boundaries (see for example [9]), they are frequently ignored by IOS design. In addition, recent works point to subtle distinctions of boundaries that are concurrently implicated through features which are either enforced by and/or inscribed into technologies [10]. Arguably, such boundaries need further attention by IOS researchers, as they offer useful insight in the direction of streamlining local/situated realities of distributed agents (i.e., the different material artifacts, norms, rules of thumb and folklore knowledge of the engaged parties). B. Boundary spanning in web 2.0 The emergence of web 2.0 has sparked a linguistic vocabulary that creates new opportunities for boundary spanning. The most typical example is currently witnessed by the use of sharing widgets and public Application Programming Interfaces (API) [11]. Sharing widgets are components that offer a way to present content directly on and across social websites or popular destinations like social networking services, thus making it available to a wider audience. They can also be used to mash-up
data feeds from different sources or across a particular topic, thus supporting a kind of interoperability. This type of boundary spanning is in sharp contrast to other forms, such as cross posting [12] that feature thematic or content boundaries which are continuously negotiated by participants. Social tagging serves as another form of boundary spanning tactic. Tags are short qualifiers (or meta-data) assigned to resources by end users. Users can freely choose tags to externalize opinion or qualify resources such as documents, pictures, music files, etc. One key advantage of tags is that they introduce an emergent shared taxonomy of keywords or classifiers, which aid users in browsing by providing meta-data about the contents of uploaded materials. Based on the notion of tags, collaborative tagging (CT) systems allow users to share resources in the web and to annotate them with freely chosen tags [13]. Consequently, CT systems are social data repositories with an embedded capacity for information and social connectivity [14][15]. Such connectivity stems from the fact that a certain user (Ui) anchors certain resource (Rj) using a tag (Tk). Then triples of the type can be used to assess a variety of constructs such as user profiles (not necessarily personal) based on tags contributed by users, resource profiles based on tags assigned to resources, recommendations, etc. More interesting for our purposes is the capacity of CT systems to facilitate enactment of cyber-structures such as virtual teams and online communities in inter-organizational settings. C. Research focus With the advent and increasing popularity of cloud-based services, IOS can extend their capabilities beyond the levels of eoperations and e-transactions towards e-Collaboration. Virtual storage management and file sharing can provide for shared resources. Social tagging components can be used to augment the value of resources shared across team members or entire organizations. Furthermore, though the use of public APIs, digital trace data analysis [16] offers new opportunities for understanding enacted cyber-structures, such as cross-functional teams, communities of practice, inter-organizational partnerships and alliances or other forms of value-based networks. The present research explores these issues in detail and presents an effort aiming to appropriate their combined benefits in an interorganizational setting. III.
ARCHITECTURAL PATTERN
Motivated by attempts to advance business connectivity in inter-organizational settings [17] and aiming to disentangle novel boundary spanning tactics [10], we have developed a system which integrates file sharing and collaborative tagging in an enterprise portal acting as the IOS. Figure 1 depicts the architectural pattern of the system and conveys the key concepts. At present our interest is concentrated on digital resources retained across (potentially) different digital spaces, such as file sharing services, social web sites or networking platforms, providing the shared data stores and „glue‟ for (intra- or inter-
organizational) virtual teams. A pre-requisite is that access to these resources is granted through the provisions made by their public APIs. Then social tagging, filtering and rating can be implemented as functions of a separate constituent – an enterprise portal, in our case – that establishes the IOS. The basic model used is an ordered tuple where Ui, Rj and Tk stand for a certain user, a certain resource and a certain tag. Users can define tags freely, although text sensitive promoting is also facilitated. Resource filtering and rating are implemented in a similar manner allowing users to augment the value of resources by anchoring them using filters and/or to externalize subjective opinion by rating the quality of an item.
Figure 2: XML version of tagging
C. Digital trace data reposiory Digital trace data is the term used to coin the capability of the system to capture data during operation and turn them into meaningful information. For our purposes, such meaningful information points out „hidden‟ patterns of use (i.e., interest relationships, social ties, tie strength, etc.) that provide evidence of enacted cyber-formations such as virtual teams. For this purpose, we use explorative knowledge visualization techniques implemented with the prefuse library (http://prefuse.org/) which is part of a tool dedicated for digital trace data analysis [19].
Figure 1: Architecture
A. Data access The Data Access Layer (DAL) is responsible for the interoperation with external digital spaces by utilizing their public APIs [11]. Consequently, DAL is constrained solely by the capabilities offered by these APIs. Depending on the chosen digital spaces, DAL may request information on the logical structure of digital resources but also meta-data (if supported) by issuing appropriate synchronous or asynchronous API calls. Part of DAL‟s functionality is implemented as servlets appropriate for each online service. When a servlet is called from the front end, it combines the parameters of this call to construct an appropriate URL. Then using this URL, it issues HTTP requests (either GET or POST for retrieving or storing) to the respective service`s public API. Finally, the response to this request is properly transformed into JSON format and sent back to the front end. B. Data aggregation / tranformation Data aggregation and transformation consolidates the data accessed into a unified repository using an appropriate internal representation (typically XML). Thus, the collection of triples is transformed as shown in Figure 2.
IV.
USE CASE AND IMPLEMENTATION
This section provides a pictorial illustration of the system. For purposes of simplicity we refrain from describing the actual interorganizational setting or other components of the enterprise portal. Thus, our working assumption is that registered representatives of different organizations articulate a collection of digital resources which is shared through Dropbox. Figure 3 provides an overview of the system. As shown the user interface assembles data of two types - API requests (through servltes) and meta-data resident locally, while making transparent the synchronization with the designated Dropbox. A. User interface overview The middle and right portlets are the core elements for codifying collective wisdom. The left portlet organizes shared resources in pre-specified categories (i.e., documents, photos, videos, etc.) and lists either selected items with their meta-data or collections of resources satisfying designated criteria. For each shared item in the middle portlet, the user can click on the file name to invoke a new window which through API requests facilitates functions such as uploading, viewing and sharing. Alternatively, the „Info‟ postfix following a resource name allows users to qualify the resource and assign meta-data. Users can explore the digital data set using filters, own tags (see Figure 3)
or community tags (Figure 4). For instance, by comparing Figure 3 and Figure 4 it is easy to detect not only how certain resources (in the middle portlet) are tagged by the user but also how peers tag the same resources. It is also possible to filter resources by rates (Figure 5), thus compiling lists of highly rated resources. The search facility available through the middle portlet entails standard local search (Figure 6 top) and advanced search (see Figure 6 bottom) that combines API requests and local processing. Standard search is keyword-based. Advanced search allows search by resource type (document, photo), filters (i.e., dates, resource type) or tags. B. Cloud-based file sharing Our current implementation uses the Dropbox file sharing service, a leading application also with Goggle Drive, Sky drive and more. At present we have used the free service, while ongoing developments are devoted to integrating other file sharing services with a mature public APIs. The choice of Dropbox was
Figure 3: Filtering resources by „own‟ tags
Figure 4: Filtering resources by „community‟ tags
grounded on the fact that in its current version, it does not provide support for user and/or content rating, filtering or collaborative tagging. On the other hand, these are both crucial for virtual team work across organizational borders. For instance, ratings offer a mechanism for users to express opinion about designated contents or even other users. Tags can also be useful either as cognitive amplifiers or as a measure of the distribution of cognition amongst members of a virtual team or a community. Consequently, it is highly desirable to devise tactics that relax or alleviate the boundary „enforced‟ by Dropbox on the use of shared resources and its implications. For our purposes, such an „enforced‟ boundary facilitates certain affordances (i.e., resource sharing) while constraining others (i.e., qualification of resources be means of meta-data). Thus, formative claim is that interorganizational and boundary spanning collaboration can be fostered by combining the Dropbox public API and the Liferay‟s portlet technology to re-locate the „enforced‟ boundary and improve upon the value of the shared resources.
Figure 5: Filtering by rates
Figure 6: Standard (top) and advanced (bottom) search facilities
C. Meta-data Our meta-data are locally compiled using Liferay‟s portlet technology. The user can define the predetermined qualifiers so as to suit the shared pool of resources and exploit the tagging mechanism to assign tags. The tagging mechanism implements adaptive prompting to allow users to gain insight on how a specific resource is tagged before they make their own tag. Each
type of meta-data is represented as triplet and stored locally. Such meta-data can be further processed to unfold „hidden‟ aspects in the data set, including shared resource communities (i.e., resources interrelated through certain tags) or virtual teams of users who tag similarly certain resources, etc. It is important to mention that by separating the functionality that manages meta-data from the referent service (i.e., Dropbox, Google Drive, etc.) provides substantial improvements on the
boundary spanning capability inscribed in the proposed solution. Firstly, the same meta-data may be used to anchor resources distributed across different digital spaces. Secondly, the same techniques for analyzing meta-data can be applied to a designated data set irrespective of the physical residence of the actual resources. Thirdly, meta-data themselves are turned into boundary spanning devices (in the sense of sharing widgets) as they may point to digital traces across bounded systems and services. Finally, meta-data may progressively become an indispensable asset of cyber-formations such as online partnerships, virtual alliances and networks of practice. Consequently, tactics such as the one presented above not only provide new and improved boundary spanning capacity but they hold the potential to unfold new joint fields between members of cross-functional, inter-organizational and global virtual teams. V.
CONCLUDING REMARKS & FUTURE WORK
The system presented in this paper is currently integrated in the Liferay enterprise portal which serves as the IOS of a research and development consortium. The consortium is tackling issues of virtual work, e-Collaboration and distributed organizing in organic farming. As the partners had no previous ties of any form and due to lack of proximity it was important to establish mechanisms to overcome differences in focus, affiliation and strategic concerns that prohibit the establishment of common ground. Using the system briefly described earlier (as integral component of the virtual enterprise), it was made possible to incrementally compile a set of shared resources ranging from documents, photographs, product leaflets, etc., and most importantly to define an emerging corpus that conveys common ground and shared understanding, thus by-passing the implications of multiple boundaries. At present, consortium partners exploit the system to tackle the problem of establishing a common classification of quality attributes across three different organic production categories (olive oil, wine and vegetables). The challenge this time stems not only from the fact that parties involved are geographically dispersed, but also from their reliance on different practices to establish quality attributes and carry out assessments. Ongoing work concentrates on integrating cloud-based services other than file sharing (e.g. online discourse and argumentation), as well as knowledge visualizations for improved activity awareness and social translucence. ACKNOWLEDGMENT This work is part of the first author‟s Master thesis in the postgraduate program “Informatics & Multimedia‟ at the Department of Informatics Engineering, TEI of Crete. The research has been co-financed by the European Social Fund (ESF) and Greek national funds through the Program "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF) - Research Funding Program: ARCHIMEDES III "Investing in knowledge society through the ESF".
REFERENCES [1]
[2]
K. Riemer, C. Steinfield, D. Vogel, “eCollaboration: On the nature and emergence of communication and collaboration technologies,” Electronic Markets 19, pp. 181–188, 2009. V. Choudhury, “Strategic Choices in the Development of Interorganizational Information Systems,” Information Systems Research 8(1), pp. 1-24, 1997.
[3]
S. Kurnia, S., and R. B. Johnston, “The Need for a processual view of Inter-organizational systems adoption,” Journal of Strategic Information Systems 9, pp. 295-319, 2000.
[4]
M-B. Watson-Manheim, M. K. Chudoba, and K. Crowston, “Perceived discontinuities and constructed continuities in virtual work,” Information Systems Journal 22, pp. 29–52, 2012
[5]
D. Shin, D. “Distributed inter-organizational systems and innovation processes,” Internet Research 16 (5), pp. 553-572, 2006.
[6]
K. Kumar, & H.G. van Dissel, “Sustainable collaboration: managing conflict and cooperation in IOS,” MIS Quarterly, 20(3), pp. 279-300, 1996.
[7]
D. Hu, S. X. Sun, J. Leon Zhao & X. Zhao, “Strategic choices of interorganizational information systems: A network perspective,” Information Systems Frontiers 13, pp. 681–692, 2011.
[8]
Y. Huang, Y. Luo, Y. Liu, Q. Yang, “An Investigation of Interpersonal Ties in Interorganizational Exchanges in Emerging Markets: A BoundarySpanning Perspective,” Journal of Management (DOI: 10.1177/0149206313511115).
[9]
P. R. Carlile, “Transferring, Translating, and Transforming: An Integrative Framework for Managing Knowledge Across Boundaries,” Organization Science 15(5), pp. 555-568, 2004.
[10] D. Akoumianakis, “Boundary spanning tactics and “traceable” connections in cross-organizational virtual alliances,” Journal of Enterprise Information Management 27(2), pp. 197-227, 2014. [11] R. Bodle, “Regimes of sharing,” Information, Communication & Society 14(3), pp. 320–337, 2011. [12] S.B. Butler, and X. Wand, “The Cross-Purposes of Cross-Posting: Boundary Reshaping Behavior in Online Discussion Communities”, Information Systems Research, 23(3), pp. 993-1010, 2012. [13] S. A. Golder, B. A. Huberman, “Usage patterns of collaborative tagging systems,” Journal of Information Science, 32(2), pp. 198–208, 2006. [14] D. Akoumianakis, and N. Bessis, “On „inscribed‟ and „enacted‟ connectivity: A tutorial introduction,” International Journal of Virtual Communities and Social Networking, 5(2), pp. 1-10, 2013. [15] J. Albors, C. Ramos, and L.J. Hervas, “New learning network paradigms: Communities of objectives, crowdsourcing, wikis and open source,” International Journal of Information Management 28, pp. 194–202, 2008. [16] J. Howison, A. Wiggins, and K. Crowston, “Validity Issues in the Use of Social Network Analysis with Digital Trace Data,” Journal of the Association of Information Systems, 12 (12), pp. 767-797, 2011. [17] D. Akoumianakis, G. Ktistakis, G. Milolidakis, K. Michailidis, “Enacted virtual partnerships: Imbrication of representations of boundary artifacts and cloud services”. In Proceedings of the 4th International Conference on Information, Intelligence, Systems and Applications (IISA2013), 2013, Piraeus, Greece, IEEE Press [18] M. Jacovi, I.Guy, and I.Romen, “Digital Traces of Interest: Deriving Interest Relationships from Social Media Interactions”. In Proceedings of the 12th ECSCW 2011, 2011, Aarhus Denmark, Springer. [19] D. Akoumianakis, N. Karadimitriou, and I. Kafousis, “Social ties in videosharing services: The tactics of excavating virtual settlements,” International Journal of Virtual Communities & Social Networking, 5(2), 27-54, 2013.
