Principles and Patterns of Organizing Systems

We have traditionally analyzed collections of information or things using categories like libraries, museums, archives, content or knowledge management systems, and data repositories. The concept of an "organizing system" complements this categorical view with a dimensional perspective that sees these categories as sets of "design patterns" that reflect typical answers to questions about what is being organized, why, when, how much, who is doing the organizing, and how services are provided to interact with the organizing system. These dimensions frame trade-offs and constraints about the content, policies, and implementation of organizing systems. The primary goal of this course is to use these design dimensions to better understand traditional design patterns and their consequences, and to identify useful new ones.

For example, the "thingness," uniqueness, persistence, useful lifetime, mashability, and intended uses and users of the content of an organizing system jointly determine how it is implemented and operated. We will examine how these design influences intersect, and consider what alternative designs would look like if some of these content and policy choices were to change. Furthermore, in many domains the Web has become the default implementation of organizing systems interfaces, yet we don’t critically examine the implications this should have on the system itself. So we will study how Web architecture – or the architectures and constraints implied by other metamodels and architectures such as "Linked data" or "WS-* interfaces" – influence decisions about content granularity and structure, how identity and provenance are supported, the kinds of interactions and services the organizing system allows, and so on.

Toward the end of the semester we will analyze some case studies and proposed architectures for health informatics infrastructures or for exposing government data in general to test and consolidate the new concepts of the course.