Yunwen Ye, Ph.D. - US grants

ABSTRACT
SEL - 0204277
Gerhard Fischer
U of Colorad @ Boulder

Title: A Social-Technical Approach to the Evolutionary Construction of Reusable Software Component Repositories PIs: Gerhard Fischer and Yunwen Ye
Although software reuse improves both the quality and productivity of software development, systematic reuse has not yet met its expected success due to difficulties in creating and maintaining reuse repositories and enabling software developers to build new systems with components from the reuse repository. The two issues are in a deadlock: if software developers are unable to reuse, the investments on reuse cannot be justified; if companies are unwilling to invest in reuse, software developers have little to reuse. This project takes a social-technical approach to address both issues. It will develop reuse-conducive environments based on the Seeding, Evolutionary Growth, and Reseeding (SER) model to guide thedevelopment and evolution of reuse repositories. It will instantiate the conceptual framework for reuse by creating a set of tools to support each phase of the SER model: SEEDER, BROKER, ENHANCER, and RESEEDER. This research will create a new understanding of the social, cognitive, and technical issues in software reuse, and explore new approaches and develop innovative tools to support reuse. The results of the research will make a significant impact on the design of information repository systems and human-centered software development environments.

We are now entering a new phase for software development where millions of people are not only using software but also becoming involved in its development to widely varying degrees. Existing software design methodologies that focus primarily on productivity-driven systems are insufficient to cope with the emergence of situated uses and fluctuating requirements encountered by such wide and diversified user involvement. Rather, a new class of participative software systems is needed, the design of which does not end at the time of deployment. For participative systems success hinges on continued user participation, and an important goal is to achieve the "best fit" between the software system and its ever-changing context of use, problems, domains, users, and communities of users. In this project, the PI will develop a meta-design framework to address fundamental challenges and guide software developers in the design of such systems. Grounded in an assessment of existing design theories as well as the systematic analysis of successful participative software systems, this research will start with a partially articulated meta-design framework, founded on the assumption that meta-designed systems can be supported by the Seeding, Evolutionary Growth, and Reseeding (SER) process model. The PI will identify and correlate the technical and social characteristics of participative software systems that support users to collaboratively engage in the design of solutions to their own problems. The identified characteristics will be used to guide future development of the Envisionment and Discovery Collaboratory (EDC), which will be used by real users to solve complex real-world problems in different design domains. The work will be integrated with a specific major, multi-year urban planning project relating to public transportation, for which an initial collaboration among the stakeholders has already been formed. Careful and systematic assessment of design decision impacts, guided by the initial meta-design framework, on this real-world problem-solving situation will feed back into the refinement of the meta-design framework. The resulting meta-design framework will delineate a design space, define a design process, and identify a set of evaluation criteria for the creation of participative software systems.

Broader Impacts: As software systems are being increasingly woven into our daily lives and reshape the way people interact, collaborate, work, and think, requirements for software systems have become more individually differentiated and continuously change during their ongoing use. This research will create the scientific foundation for the design of participative software systems that do not have fixed requirements at any point in time, and necessitate user participation and contribution as a fundamental part of the system. The research will contribute to a better understanding of the complicated interactions of technical and social aspects essential to this challenging domain of a science of design. The project will bring together researchers from design theory, software engineering, human-computer interaction, and cognitive science to gain insight into how to put owners of problems in charge and make them independent of "high-tech scribes." The many undergraduate and graduate students who will be involved in the research activities will as part of their experience be exposed to new approaches to software design.