Software design is a cognitively challenging task. Most software design tools provide support for editing, viewing, storing, sharing, and transforming designs, but lack support for the essential and difficult cognitive tasks facing designers. These cognitive tasks include decision making, decision ordering, and task-specific design understanding. To date, software design tools have not included features that specifically address key cognitive needs of designers, in part, because there has been no practical method for developing and evaluating these features. This report contains a practical description of several cognitive theories relevant to software design, a method for devising cognitive support features based on these theories, a basket of cognitive support features that are demonstrated in the context of a usable sotware design tool called Argo/UML, and a reusable infrastructure for building similar features into other design tools. Argo/UML is an object-oriented design tool that includes several novel features that address the identified cognitive needs of software designers. Each feature is explained with respect to the cognitive theories that inspired it and the set of features is evaluated with a combination of heuristic and empirical techniques.

Design critiquing systems are a type of intelligent user interface used to support human designers in decision making. This paper places design critics in the larger context of intelligent user interface approaches and surveys several critiquing systems. Each approach and system is evaluated with respect to a five-phase design improvement process. This paper concludes with a summary of the state of the art in critiquing systems and recommendations for future research directions.

Software architectures evolve as the result of numerous, interrelated design decisions. Existing approaches to analysis, however, tend to provide feedback only after numerous design decisions have been made. As a result, theydo not directly support the evolutionary nature of the architecture design process or the software architect's decision-making process. In this paper we present an approach to architectural analysis stemming from previous work in domain oriented design environments that is based on critics and criticism control mechanisms. This approach more closely supports evolution and the needs of architects by providing feedback as individual design decisions are beingconsidered. We discuss the theoretical motivations for the critic-based approach, the implementation and management of critics, support for diverse and extensible groups of critics, and the combined use of critics and existing analysis techniques.

Involving end users in the development of interactive systems increases the likelihood those systems will be useful and usable. User involvement, however, is both time and resource intensive. Intemet-and World-Wide-Web-based software release models have magnified these problems. At the same time, these practices have begun to blur the distinction between development and use and, in so doing, have provided developers with unprecedented, and currently underutilized, opportunities for increasing user involvement. We propose an approach — based on expectation-driven event monitoring and expectation agents — that leverages these opportunities to support ongoing user involvement in the software development process.

Software architecture descriptions are high-level models of software systems. Some researchers have proposed special-purpose architectural notations that have a great deal of expressive power but are not well integrated with common development methods. Others have used mainstream development methods that are accessible to developers, but lack semantics needed for extensive analysis. We describe an approach to combining the advantages of these two ways of modeling architectures. We present two examples of extending UML, an emerging standard design notation, for use with two architecture description languages, C2 and Wright. Our approach suggests a practical strategy for bringing architectural modeling into wider use, namely by incorporating substantial elements of architectural models into a standard design method.