Research And Implementation On Self-Adaptive Management And Recommendation Strategy In Mobile Collaboration Environment

Mobile collaboration has been an important project in the research field of CSCW. The aim of research on collaboration is to provide a convenient CWE (Collaborative Work Environment) for geographically dispersed groups to complete a common task. CWE provides a virtual face-to-face service by simulation. With the development of wireless communication technology, Transition is been made from traditional collaborative environment to MCWE (Mobile Collaborative Work Environment) gradually. On the other hand, with the development of context awareness, simple HCI (Human Computer Interaction) couldn't meet users who want more humanized service such as recommendation. So service of recommendation gets extensive use. The characteristics of mobile collaboration are mobility, discreteness and dynamics. So the organization in collaboration is easy to be affected by external factors. And then it may impact effect and even make the results invalid.Based on above analysis, traditional management and recommendation strategy in collaboration can't adapt to this change. It has some limitations:1) Traditional collaborative management strategy can't deal with the effect in MCE and lacks of deep analysis in the changes of management affecting the task division, coordination and intent deduce. So it lacks of effective strategies of self-adapting organization management to eliminate the effects of mobility, discreteness and dynamics.2) Recommendation service can't be provided in traditional collaboration and lack of deep study and effective self-adapting strategy.Based on above challenges, this paper studies the strategies of management and group recommendation in MCE. The main contributions can be summarized as follows:◆Theory aspect1) We analyze the features of MCWE and the effect of the features in terms of organization, and then analyze the effect in terms of recommendation.2) We propose GROUP model in two dimensional environments, defining some concepts and introducing the characteristics of it.3) We propose self-adapting management strategy based on the GROUP model, and "Multi-dimensional Weighted Voting Algorithm", which solve the leadership challenge in collaboration and eliminate the effects of leader changes.4) We propose self-adapting recommendation strategy based on the GROUP model.◆Realization aspectBased on above research methods, this paper extends related functions in LaMOC system and then presents the implementation and analyzes it with the application scenario. The result shows that the strategies in this paper make up the limitations in current research. It studies the management and humanized recommendation service, laying a foundation for future research.