برآورد اعتماد در یک تیم مجازی : روش پشتیبانی تصمیم گیری

Trust plays an important role in the operation and management of virtual teams. It is also a foundation of the cooperation between members. In the operation of a virtual team, trust estimation is a very significant issue to know about the trustworthiness of each member and cooperation situations between members. The purpose of this paper is to propose a decision support method for estimating the trust level. A trust estimation framework with two dimensions of reputation and collaboration is developed. From this, a fuzzy multiattribute decision analysis (MADA) approach based on the 2-tuple fuzzy linguistic representation model is proposed to measure the performances on reputation and collaboration. By the measurements of the performances, the trust level of each member and the virtual team can be identified so as to improve the management of the virtual team. Furthermore, a Web-based trust estimation system (WTES) is designed and developed to support the activities of trust estimation in a virtual team. Finally, an application is provided to illustrate the applicability of the proposed decision support method.

With the development of information and telecommunication technologies, more and more organizations, especially those tackling with knowledge-intensive R&D projects, are adopting the form of virtual teams to accomplish various tasks (Chen et al., 2008, Hung et al., 2004 and Kanawattanachai and Yoo, 2002). Virtual team is a group that consists of geographically and/or temporally dispersed members (Jarvenpaa and Leidner, 1999, Powell et al., 2004 and Zhu and Zhou, 2006). It has quite a number of advantages over traditional teams, such as the better utilization of distributed human resources without physical relocation of employees (Lipnack & Stamps, 1997), remote asynchronous and real-time collaboration (Warkentin, Sayeed, & Hightower, 1997) and so on. In a virtual team, members can share their knowledge and resources to jointly develop new products, new technologies, or practical solutions (DeSanctis and Monge, 1999 and Lavrač et al., 2007). However, the operation of a virtual team is more complex than that of a traditional one. Virtual team is a loose formation based on common interests of members who do not usually collaborate based on formal contracts and obligations (Lavrač et al., 2007) and lack the face-to-face communication (Cascio, 2000, Hung et al., 2004 and Jarvenpaa and Leidner, 1999). In the operation and management of a virtual team, trust plays a very important role (Greenberg et al., 2007, Kanawattanachai and Yoo, 2002, Lipnack and Stamps, 1997 and Piccoli and Ives, 2003) and it is also a foundation of the cooperation between members (Kanawattanachai and Yoo, 2002 and McAllister, 1995). The higher trust level of members could produce more positive attitudes (Dirks & Ferrin, 2001), maintain favorable cooperation relationships (Zhang, Liu, & Zhu, 2008), facilitate communication and knowledge sharing between members (Levin and Cross, 2004 and Renzl, 2008), and finally promote the team performance (Iacono and Weisband, 1997, Jarvenpaa and Leidner, 1999, Kanawattanachai and Yoo, 2002 and Keyzerman, 2003). Whereas, the lower trust level of members would hamper a team to accomplish tasks effectively (Iacono & Weisband, 1997). Thus, it is a necessary and significant work to estimate the trust level of members in a virtual team. Through trust estimation, team leader(s) could know about the trustworthiness of each member and cooperation situations between members. Some studies dealing with the trust estimation problems can be found (Castelfranchi and Falcone, 2001, Chen and Chen, 2009, Chen et al., 2008, Lavrač et al., 2007, Msanjila and Afsarmanesh, 2008, Mun et al., 2009, Schmidt et al., 2007 and Zhang et al., 2008). In these studies, reputation and collaboration are regarded as two important perspectives of trust estimation. Furthermore, the trust level is usually measured by aggregating the performances on reputation and collaboration. For example, Lavrač et al. (2007) proposed a hierarchical multiattribute decision support model to estimate the trust level of a collaborative network organization, where the trust level is measured by computing a weighted sum of scores of reputation and collaboration. Various trust estimation models proposed in previous studies significantly contribute to solving the problem of trust estimation. However, some limitations occur when these models are applied. Although the trust level can be estimated using a model by aggregating the performances on reputation and collaboration, e.g., a weighted sum model, the trust estimation result could not clearly reflect the real situation of reputation and collaboration, respectively. In deed, reputation and collaboration should be respectively measured because their meanings are different. If the performances on reputation and collaboration are both higher, then the trust estimation result could really reflect the higher level. On the other hand, few models for estimating the trust level involve processing multiple formats of assessment information, such as fuzzy linguistic terms, crisp numbers and so on. In fact, both subjective assessment information with human perception and objective assessment information based on actual situations are often used in trust estimation, where they can be of multiple formats. Moreover, it is necessary to point out that the existing studies on trust estimation mostly focus on models or methods, whereas implementations of decision support systems for trust estimation are neglected. It is also important how to design or develop a decision support system since the trust estimation in a virtual team is a periodic activity that involves several decision procedures. Hence, to solve the above problems discussed, it is necessary to develop a new decision support method for estimating the trust level. This is the motivation of our study. The purpose of this paper is to propose a decision support method for estimating the trust level. A trust estimation framework with two dimensions of reputation and collaboration is developed. On the basis of the framework, a fuzzy MADA approach is proposed to measure the performances on reputation and collaboration, where the 2-tuple fuzzy linguistic representation model (Herrera & Martínez, 2000) is used to process fuzzy linguistic terms and crisp numbers involved in the estimation process. Furthermore, to solve the practical trust estimation problems using the proposed method, a Web-based trust estimation system (WTES) is designed and developed to support the activities of trust estimation. This paper is organized as follows. Section 2 presents the background and related work on trust estimation. Section 3 presents a trust estimation framework with two dimensions of reputation and collaboration. In Section 4, a fuzzy MADA approach based on the 2-tuple fuzzy linguistic representation model is proposed to measure the performances on reputation and collaboration. In Section 5, a WTES is designed and developed to support the activities of trust estimation in a virtual team. In Section 6, a real example is used to illustrate the application of the proposed method. Finally, Section 7 summarizes and highlights the main features of the method proposed in this paper.

This paper proposes a decision support method for estimating the trust level in a virtual team. In the proposed method, a trust estimation framework with the two dimensions of reputation and collaboration is presented. Then, a fuzzy MADA approach based on the 2-tuple fuzzy linguistic representation model is proposed to measure the performances on reputation and collaboration. Furthermore, a WTES is designed and developed to support the activities of trust estimation in virtual teams. The major benefits of the proposed decision support method are: • The trust estimation framework provides a clear procedure for estimating the trust level of a virtual team. The framework consists of the two dimensions of reputation and collaboration. Reputation and collaboration are respectively measured to identify the trust level of team members, or of a team. • The fuzzy MADA approach is presented to measure the performances on reputation and collaboration under fuzzy linguistic environment. The approach is based on the 2-tuple fuzzy linguistic representation model. It has clear logic and simple computation process. It has the advantage that there is no information loss when fuzzy linguistic terms are processed. • The developed WTES can be conveniently used to support the distributed members of a virtual team to participate in the activities of trust estimation. It can support the team leader to obtain the estimation results on the two dimensions automatically. According to the obtained results, the team leader could give the targeted measures to improve the trust level. The proposed method has been applied to estimate the trust level of PPCRT team. It can also be extended to support the trust estimation in other situations, such as new product development teams (Fan et al., 2009 and Kim and Kim, 2009), cross-functional teams (Feng et al., 2010 and Wang et al., 2003) and so on.