Monday, March 13, 2006
Strawman definition and Open XMP
Here is my first cut at a strawman definition:
● Metadata – an external description of a distinct data resource to provide context, metrics or amplification.
Let me know what you think...
On another note, Adobe announced Open XMP. I am looking forward to checking this out. It is good to see Adobe take a leadership role in this space. Note: XMP uses the W3C RDF specification.
● Metadata – an external description of a distinct data resource to provide context, metrics or amplification.
Let me know what you think...
On another note, Adobe announced Open XMP. I am looking forward to checking this out. It is good to see Adobe take a leadership role in this space. Note: XMP uses the W3C RDF specification.
Comments:
<< Home
In general, I think most metadata definitions are problematic because they don't start with an attempt to tease out why people are using metadata. Good definitions are grounded in a sense of why people need the concept -- and metadata is hard to define precisely because that's all over the map. Here's my take on a definition:
"metadata" is a term that refers to many different classes of information that provide an indirectly useful description of some data asset. "Metadata" may refer to: (now create sub definitions for resource metadata, context metadata, etc)
The "indirectly useful" part is important (albeit admittedly a little geeky) to emphasize that metadata is not the content itself - it augments the content. In other words, metadata isn't directly used in support of the purpose as data is.
But since you're looking for feedback on this particular definition, I can mention a few points.
- Why emphasize an "external" description? What about embedded tagging?
- Why say a "distinct" data resource? If I have information that provides context to a class of information resources (i.e. all RDBMS share some common schema) - then it's not metadata because it doesn't refer to a distinct data resource?
- Saying "context, metrics, or amplification" may not work. There are countless use cases, and approaching it as a list (i.e. metatadata provides X, Y, Z, or A) will be incomplete no matter how long the list.
"metadata" is a term that refers to many different classes of information that provide an indirectly useful description of some data asset. "Metadata" may refer to: (now create sub definitions for resource metadata, context metadata, etc)
The "indirectly useful" part is important (albeit admittedly a little geeky) to emphasize that metadata is not the content itself - it augments the content. In other words, metadata isn't directly used in support of the purpose as data is.
But since you're looking for feedback on this particular definition, I can mention a few points.
- Why emphasize an "external" description? What about embedded tagging?
- Why say a "distinct" data resource? If I have information that provides context to a class of information resources (i.e. all RDBMS share some common schema) - then it's not metadata because it doesn't refer to a distinct data resource?
- Saying "context, metrics, or amplification" may not work. There are countless use cases, and approaching it as a list (i.e. metatadata provides X, Y, Z, or A) will be incomplete no matter how long the list.
# posted by 
: 10:05 PM
: 10:05 PM
Thanks for your insightful comments. I need to think about what you have said but here are some answers to a few of your questions:
1. Why "external description"? My thought here is that a key part of the problem is blurring between the definition of metadata and the definition of data. Thus, I am proposing a key distinction being that metadata is separate and distinct from the data it describes.
2. See above for "why distinct"?
3. You are absolutely right on the laundry list. I had made the same comment on Craig Tanner's definition (see previous comments) and then turned right around and made the same mistake. The one difference I strived for was to attempt higher-level categories that provide the key purpose of metadata (which was my attempt to address your initial point about why people would use metadata).
On your indirectly useful aspect ... very interesting. It would be good to see if we could quantify that.
Regards,
- Mike
Post a Comment
1. Why "external description"? My thought here is that a key part of the problem is blurring between the definition of metadata and the definition of data. Thus, I am proposing a key distinction being that metadata is separate and distinct from the data it describes.
2. See above for "why distinct"?
3. You are absolutely right on the laundry list. I had made the same comment on Craig Tanner's definition (see previous comments) and then turned right around and made the same mistake. The one difference I strived for was to attempt higher-level categories that provide the key purpose of metadata (which was my attempt to address your initial point about why people would use metadata).
On your indirectly useful aspect ... very interesting. It would be good to see if we could quantify that.
Regards,
- Mike
<< Home
