3. Descriptive Metadata
3.1 Assess Current Metadata
Before beginning, it is important to profile the data that will be used. One must know what data is available for use in building the interface before developing specs. Certain attributes may affect what is possible within the search interface, or may require pre-processing of the data. If the data does not yet exist, the following are things about which to consult before data entry begins. The data should be profiled on the Metadata Aassesment Tool.
3.1.1 System(s) and format(s) in which metadata currently exists
- System e.g.,
- Horizon
- Online finding aid
- Stand alone database or application (e.g. MSAccess, MSExcel,MySQL, etc.)
- Paper
- Format e.g.,
- MARC
- Dublin Core
- EAD
- Library-created database
- Other
3.1.2 How accessible is the metadata now?
- Available to patrons onsite?
- Available globally online?
- Available to staff only?
3.1.3 Level (or granularity) of exisiting metadata
- collection-level?
- series-level?
- item-level?
3.1.4 Status of records
- Quality of record (recon? fully cataloged? etc.)
- Scheduled for conversion/upgrade?
3.1.5 Key features of the data
It is especially important to describe the key features of the
data if the data currently exists in a non-standard format. Data in
standard formats will ususally have more predicable features
because of the known requirements of the standard and the system in
which it exists. Records should still be examined for
anomolies.
Examples of things to look for include:
- Characters
- Accents, foreign characters, and symbols
- Punctuation included in fields
- Symbols within fields which have a special meaning
- Data structure
- how cleanly differentiated is data; how well does content in fields match the field definition
- How is data entered (e.g., separate fields for first and last name? last, first in one field? first last in one field?
- Has authority control been used?
- Are there blank fields (does every field hold a value or not)?
3.2 Project Metadata Decisions
3.2.1 Metadata system(s) and accessibility
- What systems should project metadata reside in? at what granularity?
- What other systems should the metadata be able to interact with?
3.2.2 Data elements
A complete list may emerge as work is undertaken and as interface specs are developed. It is useful to begin a list with what is known to be needed and continue to add to it throughout the project.
3.2.3 Processing of existing metadata
What work needs to be done on exisiting metadata in order to utilize it in the project?
- data entry
- data clean up
- adding new metadata
- data transformation