The subject element is used to describe the subject content of a digital resource. It is also used to facilitate searching and browsing a digital collection by subject and collocating resources with the same subject content.
The subject element is important in a descriptive metadata schema because of its role in the functionality of searching a digital collection. When users are seeking content in a digital repository, they will likely do exploratory searching or browsing by subject. All metadata schemas have a subject element for that reason. Using a controlled vocabulary is recommended for subject element values to aid discovery. When creating subject element values, the subject content needs to be represented by words that can be retrieved.
Prior to entering the controlled vocabulary terms for subject content, the subject content of the resource needs to be analyzed. This can make the metadata subject values prone to subjectivity and ambiguity. Non-textual digital resources are especially dependent on the metadata creator’s interpretation as there are no words to search for content. The interpretation of the subject content can present some challenges: how much content description to include; what the subject content is of versus what it is about; the amount of time allotted to researching; determining the subject content important to the particular digital collection.
Steven Miller suggests the following steps in analyzing subject content of images:
1. Consider the mission or main topic of the digital collection and the users’ needs.
2. Determine the main focus of the image. Can you glean any information from a supplied title or notes?
3. Formulate subject concepts into concrete terms—most likely using established controlled vocabularies.
4. When entering subject terms, ask—what will be the result of users’ search?
The information in the textbook has given me a lot to consider when applying values to the subject element of our metadata schema. It seems that a good place to start would be identifying the focus of the digital collection and the likely users’ needs. Understanding the search and browse functionality will be key. Incorporating a controlled vocabulary such as the Getty AAT or the LCTGM will be important as well. I plan to check the Getty Guide to the AAT, the Intro to the LCTGM, and Using Dublin Core next.
Source:
Metadata for Digital Collections, by Steven J. Miller
Michele Quinn 
Great summary of the subject element! I thought Steve Miller’s steps for analyzing subject content for images will be a helpful focus for us in the indexing project.