IDENTIFYING, RELATING, AND EVALUATING DESIGN PATTERNS FOR THE DESIGN OF SOFTWARE FOR SYNCHRONOUS COLLABRATION.

Many working environments require that geographically distributed or colocated work group members work together - supported by software - in developing and refining one commonly shared resource in the same time. In Computer-Supported Collaborative Work, this is defined as synchronous collaboration. Domains subject to such type of collaboration are many, some such examples are drawing, searching, text editing, or game solving. Technology has helped switching from the real to the virtual, simplifying such collaborative efforts and providing technological support for more efficient and faster collaborations. Several software applications, developed as either research projects or commercial products, exist and are used today in synchronous collaborative settings in domains such as drawing, searching, text editing, and game solving. In order to further support such developments, there is a growing need for knowledge capturing and sharing with respect to the challenges and the concerns to be faced in designing such tools. Surely, documentations of existing applications help, but it is only a larger and perhaps more general repository of knowledge that would do a better job. Several approaches for building such repositories of knowledge exist, the approach this thesis is further exploring is design patterns. On one hand, I aim at bringing methodological support to design pattern research in answer to the scarce landscape of methods and techniques for both identifying design patterns in interaction design and generating pattern languages based on existing collections of patterns. I focus mainly on one area of interaction design, i.e. the design of applications addressing synchronous collaboration, and I target four domains in the area, i.e. drawing, text editing, searching, and game solving. On the other hand, I am interested in better understanding how design patterns are used and what is the impact of using them in collaborative design processes. I first focus on the collaborative processes involving novice designers, aiming to correlate the findings from this initial study with those obtained after investigating similar processes involving experienced users of patterns. This work primarily impacts design pattern research at a methodological, theoretical, practical and empirical level. Secondarily, the findings described throughout the thesis inform behavioural research and humancomputer interaction. At a methodological level, I describe two methods addressing design pattern research; one is used for identifying design patterns in interaction design, while the second one is used for generating pattern languages out of existing collections of patterns. At a theoretical level, I describe the results of applying the pattern identification method in the area of synchronous collaboration, providing 15 design patterns addressing the design of applications targeting this area. Practically, I developed a software application able to support the semi-automation of the pattern language generation method and the execution of queries on the output of this generation process. At the empirical level, I present a case study designed to bring some light into the matter of collaborative use of design patterns. The results of this case study aim at identifying strategies novice designers develop while collaboratively using design patterns during interaction design processes.