This paper discusses VIEWS, a specification for building diagrams that describe the security features of systems. The authors' recent experience with providing security architecture and engineering support to organizations with large, distributed applications suggests that security architecture and assurance efforts could benefit by following other engineering disciplines, where using graphical models is the norm. Security diagrams can help security architects understand a system 's security posture and can assist them in detecting vulnerabilities. Additionally, diagrams facilitate communications about the security features of a design. The output of a modeling effort using VIEWS is a diagram depicting a system's security features as well as those of the environment in which the system operates. A goal of VIEWS is to allow the display of important security features without injecting cluttering detail. This paper presents examples of security diagrams built with VIEWS.

Shared scientific data, such as neuroimagery, offers great benefits to science. However, data owners must exercise custodial responsibilities which can conflict with the unhindered sharing of their data. Given simple choices of sharing widely or not at all, the result will frequently be no sharing. We hypothesize that neuroimagery sharing will be enhanced if data owners are provided with well-defined intermediate levels of data visibility. In this paper, we describe a broadly applicable data sharing model, Structured Sharing Communities (SSC), in which data becomes incrementally visible to communities structured as a complete partial-order; the associated properties of Privacy and Fairness regulate access to shared data. Within SSC, a customized policy space is defined capturing the sharing relationships among specific collaborators.

This paper focuses on the problem of management of CAD data by investigating the use of the software engineering principles of data abstraction and object-oriented programming as means for managing the volume and complexity of design data. It discusses how a data management system which manages abstract data types provides a framework for integrating a CAD environment by providing a single mechanism to manage data, programs, and the manipulation of the data with the programs. It also shows how two forms of abstraction, aggregation and generalization, can be used to model hierarchies of parts and components, and multiple design alternatives, respectively. These abstractions are then used to model the geometrical and topological data required to describe three-dimensional mechanical parts.