Metadata Understanding
Started by Daly de Gagne
on 1/24/2007
Daly de Gagne
1/24/2007 12:31 am
In keeping with the previous thread about using metadata in UltraRecall, I am curious to know how many of us have a real sense of how we can use metadata to advantage. I am not sure I have a full understanding, although I have used metadata in various forms in InControl, a wonderful old Mac program, in Claris Organizer, ADM, UR, and MyInfo.
It seems to me it takes a certain way of thinking or visualizing one's information to see how best to use metadata features. I am curious about the different ways people here use metadata.
Daly
It seems to me it takes a certain way of thinking or visualizing one's information to see how best to use metadata features. I am curious about the different ways people here use metadata.
Daly
David Dunham
1/24/2007 3:00 am
I have used metadata in various forms in InControl, a wonderful old Mac program
InControl had metadata? Maybe the copy on my shelf is too old (and I don't even know if I can check since there's probably a floppy disk in the box). I know it had columns.
I'm particularly interested since to some degree I'm modeling parts of Opal on InControl.
Daly de Gagne
1/24/2007 4:13 am
The columns of InControl could contain metadata -- I loved that program with a passion, and felt sick when I learned it wasn't being developed anymore.
Daly
David Dunham wrote:
Daly
David Dunham wrote:
>I have used metadata in various forms in InControl, a wonderful old Mac
program
InControl had metadata? Maybe the copy on my shelf is too old (and I don't
even know if I can check since there's probably a floppy disk in the box). I know it had
columns.
I'm particularly interested since to some degree I'm modeling parts of
Opal on InControl.
Graham Rhind
1/24/2007 7:47 am
I think that if you wish to discuss metadata you need always to define what you understand with the term as it is overused and too often unspecified. If metadata is "data about data" or "structured, encoded data that describe characteristics of information-bearing entities to aid in the identification, discovery, assessment, and management of the described entities"; then any outliner contains metadata, because a note name is metadata, a keyword is metadata, a title is metadata. Even the words within the text could be regarded as metadata because they can be searched.
What do you understand by "metadata"?
Graham
PS - I never use the term myself, and I'm a data professional ;-)
What do you understand by "metadata"?
Graham
PS - I never use the term myself, and I'm a data professional ;-)
Ken Ashworth
1/24/2007 11:36 am
Graham Rhind wrote:
...then any outliner contains
metadata, because a note name is metadata, a keyword is metadata, a title is metadata.
Even the words within the text could be regarded as metadata because they can be
searched.
Yes, all of these programs overlay a database, or in it's simpilist terms (abstract) a two-dimensional table - where each row is a record (node or item), and each column is a Field with the intersecting cell containing the FieldData. Starting with a new database, or empty table, you populate the table by adding nodes.
Since each node is a record, everything that describes (defines) the node is an attribute (FieldName:FieldData or column).
All programs reveal a basic set of node attributes - Node Title and display characteristics (font apperance), node detail (user entered data). Some reveal, and make available, a set of system fields - default fields required for each node.
For example: the inclusion of a time/date stamp in the node detail is usage of a system field, or changing the icon of a node in the tree pane. The availability (to the user) of these system fields varies among programs.
For example: Treepad allows for the placement of such system fields as CurrentNodeTile, ParentNodeTitle, Sibling... within the item detail (RTF editor).
Taking this to the next level are those programs that allow for user-defined attributes (FieldName:FieldData or column).
Continuing with the example of UR, starting from a new database select a node, then Tools|Attributes, then click on the column heading Category to sort the list. Now you can scroll thru the list and see those fields that are considered System Fields (checked) and those that are User-Defined (not checked).
From Tools|Attributes window you can select and Insert an existing field and it will be revealed in the Item Attributes pane (View|Other Windows|Item Attributes), or you can create a new (user-defined) field, then select and Insert.
Through the Tools|Attributes window and the Items Attributes pane, UR reveals (and makes available, data entry) the coulmns of our two-dimensional table.
Since each record (row or node) contains all columns, all columns are available for each node. It just depends on which column (field) you want to reveal in the Item Attributes pane.
Those programs that allow for the creation of user-defined fields add a level of structure to the freeform nature of these programs. In the case of UR you can add further structure thru the use of Templates.
Later,
KenA
Stephen Zeoli
1/24/2007 11:54 pm
Daly,
When I think of meta data, what I think of is information that can be attached to foundational information. In a single-pane outliner, the foundational information would be the node itself. In a two-pane, or three-pane outliner, the foundational information would be the data stored in the editor window. The meta data in each case would be additional information associated with this foundational information to help organize and analyze the outline's contents.
GrandView did this beautifully, allowing you to create custome tags for an outline. So you could build a to do list for a project where each node would be a milestone or task, and you could create categories for priority, due date, assigned to, etc... Then, in a separate categories screen, it would show you a list of all the various entries for any one category. That is, you could select the priority category and see of list of all the prioritizations you used. For example, select "high" and it would show you all the items you categorized as high priority. Or in the due date category, you could view all the items due on a certain date. Essentially GV allowed you to "hoist" your information by category -- this is probably a lot like InfoHandler, come to think of it. You can achieve a similar result with Zoot.
Now, the meta data columns available in a program like MyInfo are not nearly as useful, even though they can be helpful.
Steve Z.
When I think of meta data, what I think of is information that can be attached to foundational information. In a single-pane outliner, the foundational information would be the node itself. In a two-pane, or three-pane outliner, the foundational information would be the data stored in the editor window. The meta data in each case would be additional information associated with this foundational information to help organize and analyze the outline's contents.
GrandView did this beautifully, allowing you to create custome tags for an outline. So you could build a to do list for a project where each node would be a milestone or task, and you could create categories for priority, due date, assigned to, etc... Then, in a separate categories screen, it would show you a list of all the various entries for any one category. That is, you could select the priority category and see of list of all the prioritizations you used. For example, select "high" and it would show you all the items you categorized as high priority. Or in the due date category, you could view all the items due on a certain date. Essentially GV allowed you to "hoist" your information by category -- this is probably a lot like InfoHandler, come to think of it. You can achieve a similar result with Zoot.
Now, the meta data columns available in a program like MyInfo are not nearly as useful, even though they can be helpful.
Steve Z.
