Overall Rationale: We chose to start with the general descriptors that have very wide parameters thus always having the potential to be narrowed as content drives more categorization. We also chose to describe our pictures in two ways:

Contextualized content:

Content that is placed within the physical and intellectual context that existed at the time the picture was taken.

General content:

Aimed at describing content that focuses on the actions and objects within the picture regardless of the physical and intellectual context.

Facets:

• Academic - This was probably our most difficult facet due to the interdisciplinary nature of our project. We chose to begin with the larger genre of Academic to signify some sort of Academic work is being done by all components of this facet. To incorporate the wide variety of projects and impromptu collaborations within and between departments, we divided the academic into traditional academic departments and emerging interdisciplinary topics.
  
• Actions - This facet describes general actions reflecting human (or machine) movement (i.e. walking, standing).
  
  
• Activities - This facet reflects events that are university related as well as shopping or holiday activities.
  
  
• Community- this category differentiates between various community oriented resources such as entertainment and transportation.
  
  
• Environment - Another general facet aimed at describing the general environment of the photo (i.e. landscape, weather, time of day).
  
  
• Object - The object facet allows users to search by inanimate objects within pictures regardless of the context.
  
  
• People - This is a critical facet to our design. People are central to describing ties between inanimate categories for departments, interdisciplinary topics, and activities. We divided people by Berkeley affiliation, and then by professor, student, or staff roles.
  
  
• Photo_Properties - The categories within this facet reflect elements specific to the nature and quality of the photograph itself (i.e. file format, lighting, photographer).
  
  
• Places- Another large component of our application, this facet describes the buildings and facilities within the photos. We chose to break it down by campus/non_campus to reflect the natural boundary between the campus and the outside world, and to identify those items to be mapped in our application.
  
  
• Resources - This facet is fairly narrow and may need reworking, but we found we needed something to describe the student resources on campus.
  
  
• User_permissions - While we have not yet taken any pictures that require a release, we anticipate this facet will be needed in the future as well as by other applications.

Weaknesses: We started on a very general level to ensure that our initial facets were wide enough both for our own needs and those of other applications. In order to suit our application better we may need to rework sections to ensure all of our objects are easily retrievable.

Strengths: We have worked to develop a comprehensive classification model that is focused on reusability and scalability as the nature of application grows and changes. In addition to contextualizing content, our system seeks to account for more generic descriptors that will enable users to retrieve photos based on their content.