MetaTagger Architecture Question Help

thriventman

Good Morning,

I was wondering if anyone can tell me the difference between using the MetaTagger's "Tagger Gui" interface versus the Vocabulary. Can you still use the Vocabulary without building the Tagger Gui?

The questions come from the people that I support and I don't really understand how the 2 coincides with each other. The Tagger Gui can give you an option to add customized metadata. However, the vocabulary lets you add the alternatives and clues but not the customized metadata.

The vocabulary creates the index correct? And the Tagger Gui just gives them the option to customize the metadata. The users want to know if I put an alternative in the vocabulary, how will it add that to the metadata?

They just don't want to build the Vocabulary which is the gist of the question...and see if they can just have the Tagger Gui?

Any suggestions? I hope I made sense

Thanks Everyone!

Migrateduser

Hi Thriventman,

The two UIs serve different purposes. What is affectionately referred to as the Tagger UI is the end-user/business-user UI. It performs the several functions:
1. It displays the results of the automated tagging in a form-like interface for users to review and refine
2. It enables users to browse controlled vocabularies to manually slot content in the proper category
3. It provides the capability to enter arbitrary metadata for documents by typing values into a form

The Tagger UI can be invoked via TeamSite workflow or manually launched via a TeamSite menu option.

Studio is a vocabulary development environment. It enables a specialist to model, test, and deploy your controlled vocabulary so that users of the Tagger UI can automatically/manually classify documents into topics. Also keep in mind that controlled vocabularies can be compiled as a Recognizer which identifies key terms in content that are specified in the controlled vocabulary, and can resolve any ambiguity at the same time.

Now, it's up to you whether you want to utilize controlled vocabularies (classifiers and recognizers). They're very powerful, and take a lot of burden off the user by automatically processing content with respect to topics or key terms. However, they're certainly not mandatory in all cases. For example, if you want to capture dublin core metadata, you should be able to capture these elements (author, title, description) with a high degree of automation by using some of MetaTagger's other capabilities like preprocessing, extraction, summarization, etc.

HTH.

Mike

Edited by msvatek on 09/16/03 10:21 AM (server time).