Nelson and Haupt - Rockley Chapter 9: Metadata
Metadata is “data about data.” Any item of data is a description of something. Metadata is a type of data where the something being described is data. There are two types of unified content metadata: categorization and element. Users tend to use categorization to search for metadata, yet authors will most likely us an element search to find metadata.
An example of metadata would be to recognize a date on a photograph then when later searching for that practically photograph a user could search for the photograph by the date on it.
when compiling metadata it is important to understand who is going to retrieve the content, what tasks are they trying to accomplish with the content, and what terms will they use when retrieving the content. After understanding the first three steps in creating metadata it is important to create taxonomies. A taxonomy is the grouping of like or similar content together. After you have grouped your taxonomy you will want to test it. To test the taxonomy you will want to run a usability test on it.
When working with element metadata there are three main types: reuse, retrieval, and tracking metadata. Reuse is if data already exists in a data base then the metadata will recognize and ask the user if they would rather go back and reuse the data in it. Retrieval metadata is used to help authors retrieve content and information. Examples of information in a retrieval metadata would be the title/ subject, author, date, and keywords, and the security level. Metadata for tracking is used when showing workflow.
Metadata needs to be consistent to facilitate reuse, retrieval, and tracking. To keep metadata consistent it is best if a controlled vocabulary is created.
An example of metadata would be to recognize a date on a photograph then when later searching for that practically photograph a user could search for the photograph by the date on it.
when compiling metadata it is important to understand who is going to retrieve the content, what tasks are they trying to accomplish with the content, and what terms will they use when retrieving the content. After understanding the first three steps in creating metadata it is important to create taxonomies. A taxonomy is the grouping of like or similar content together. After you have grouped your taxonomy you will want to test it. To test the taxonomy you will want to run a usability test on it.
When working with element metadata there are three main types: reuse, retrieval, and tracking metadata. Reuse is if data already exists in a data base then the metadata will recognize and ask the user if they would rather go back and reuse the data in it. Retrieval metadata is used to help authors retrieve content and information. Examples of information in a retrieval metadata would be the title/ subject, author, date, and keywords, and the security level. Metadata for tracking is used when showing workflow.
Metadata needs to be consistent to facilitate reuse, retrieval, and tracking. To keep metadata consistent it is best if a controlled vocabulary is created.
posted by Carl Haupt | 7:57 AM
8 Comments:
Well, I guess I'm not having a deja vu experience. I totally misread the assignment sheet from last week! Sorry I did Chapter 9 when I should have done Chapter 8 in Rockley. (At least I got the right textbook.) I'm printing off the reading assignment sheet. I think this is a true example of someone scanning the Web instead of reading word-for-word. Anyone want me to post Chapter 8?
As search engine users, we are most familiar with the concept of categorization, I think. Most search engines probably use categoriztion to locate the multitudes of pages when we search for something on the web. The categories might be broad or they might be narrow; it isn't very often you find a "Googlewhack" as a result of a search (only one result answer). I particularly liked the author's analogy to a library card catalogue, which might give an abstract as well as the location data.
The idea of element metadata seems to be the main focus of this author throughout this book. This type of data would make content reuse much easier by making it easier to locate, retrieve, and update anything that is part of a unified content strategy. It is more the who, what, when, where, and why of the data.
I also think that the use of a controlled vocabulary is probably very important for cosistent retrieval and reuse of data content.
To continue with Larry's supporting statement about controlled language, I think consistency of language is essential to creating a content use strategy. All users of the content reuse system should have a common understanding of the language being used. Inconsistent language use is one of the content problems Rockley addresses in the beginning of this book. When multiple authors are using multiple pieces of content that are supposed to mean the same thing, the reader will be left confused about what to believe is correct.
I can see how metadata must be essential to a content reuse system. If content is not organized, no one will be able to use the system. If content is not organized in a way that is comprehensive to the users of the reuse system, then how would they be able to find what they need?
It's amazing how different your results can be when searching for something and re wording it even slightly. To go along with what Larry and William have said I think it is very important to use consistent language in order to create an easier reuse element among documentation. It can be very frustrating when you're sure your searching with the right criteria and end up getting results that are totally opposite of what you're intending.
I guess since I just responded to the blog of Anne's Chapter 9, I won't respond here again.
Utilizing metadata in a business scene can lead to a more efficient search if the person searching does not know the specifics of the document that is needed.
I think that a consistent vocabulary would be nice but how would you go about installing one? Who gets to make it and how do you decide who/what is added to it? How do you ensure that it is being used? it seems like a really difficult thing to use.
Metadata doesn't seem so much like data about data, more like portions of a whole. You have a group of data (photograph, date it was taken, photographer, location, etc), and you can search for a specific chunk of that data to find the entry you're looking for. At first I thought you were going to talk about statistical information about data; how much there was, in how many divisions, etc. If you were a large data mining company, you might need another layer of statistical analysis just to put in simple terms the vast amount of metadata you have! It's like an informational echer sketch.
Post a Comment
<< Home