FirmTagMethedology

From Datafedwiki

(Redirected from TagMethedology)
Jump to: navigation, search

There are several different kinds of content on the wiki. I decided based on the structure already in the wiki and the DataFed website that the content is:

  • Datasets
  • Events - A place for AQ Events
  • Reports/Proposals
  • Projects
  • WebApps - combined in the wiki from applications and tools since in DataFed it was just one tag
  • WebServices
  • DataFedWiki - I just thought of, but I have been tagging wiki related content with this tag.

Besides for different content wiki pages also have a page status (done/inprogress/new), page structure (atomic/composite), possibly a date and/or location.

Datasets already had the strongest structure imposed because of the catalog and some of the datasets had already been semi-tagged in the wiki. I took these pages and added the rest of the catalog tags for Format, Method, Provider/"Belongs To", Platform and Domain. I then tagged the domain tags I had used in the datasets - aerosol, fire, meteor etc. as domain and I tagged domain with Datasets. I also did this with the other types of catalog tags. Because of the standardization of these tags a hierarchy can be imposed, but it has many axis Datasets -> Format -> Aerosol-> SeaWifs_US or Datasets -> Platform ->Satellite->SeaWifs_US will get you the same place. Datasets also get page status tags if they aren't finsished and page structure tags (atomic). Clicking on the catalog page you can see the pages filtered by domain.

CATT was another example I used to base the new tagging navigation. CATT was a WebApp and Project so the main page has those tags. It was working on CATT that I realized the different kinds of content could have multiple parts. The parts are generally standard. For example CATT had bugs and wishes, but the template for bugs for CATT isn't very different from the bugs for the Viewer or Datasets - In fact it is the addition of a Dataset tag or a Viewer Tag that changes the bug template. So I came up with a list of page types (bugs, wishes, about/main pages, news, todo, collaboration, conventions) not each type of content will have all of these but, by having the same general template it means that we can have a list of wishes compiled over the whole wiki or it can be filtered by "belongs to" or content type. WebApps -> CATT and then CATT has multiple pages underneath it for wishes, help, bugs etc.

These firm tags need to have a prefix which identifies the class of tag to make it unique. Examples: Dom.Aerosol Page.Done, Page.Atomic, Date.yymmdd etc. This way the loose tag aerosol won't be mixed up with the domain specific aerosol. We also need to sharpen the DPL filters to search for part of a tag i.e. first letters = date.05 or geo.lat in order to search for date range/ location range.

Personal tools
Workspaces
Clicky Web Analytics