Triple-Tagging

Help Topics

As shown on the Welcome Page, Triple-Tags have:

a subject, a property and an object value

A complete triple-tag consists of web addresses for each part of the triple. Our tool lets you construct these triples easily. You can create as many triple-tags for an image as you would like. There can be multiple triple-tags with either the image or any region as the subject.

The triple-editor automatically searches datasets to find appropriate web address 'entities' so you can specifically define your object value/keyword and create highly useful triples for search engines.

Double-click on any image thumbnail displayed in the "My Images" window to open the metadata editor and get started.

To Add Triple-Tags:

This image illustrates two ways to create triple-tags.

You can click the 'Add' button, to add a triple-tag at any time.

Using the Region Creation Tools will also let you add a triple-tag as you create the region.

The first triple-tag in the editor always defaults to the depicts property, but the creation of all other triple-tags will automatically drop-down a list of recommended properties. There are no 'rules' to using these properties...you cannot do it wrong, but reviewing the Visual Guide and doing some sample searches will help you see how their use can give you an added advantage in searches. Advanced Users can use other properties. Please see: Ontology Use for more information.

When you have entered a word or phrase for your object value/keyword, the characters will first display in red and a black exclamation point will appear.

The red text indicates that a correct definition has not yet been entered for the term and the black exclamation point indicates that a possible match has been made to an entity in a datasets used for the look-up.

Select the correct definition for the term by clicking on the black exclamation point that appears.

If a red exclamation point appears, it indicates that an appropriate match was not found, but you can still select the exclamation point to create a new entity - see Creating New Entities and Datasets for more information

Note: Creating the most useful triple-tags should be accomplished by always choosing or creating a correct entity for a term. However, there are cases where an error was made and a definition was not selected or where the user deliberately chose not to select an appropriate entity. In this case, the triple is still stored with the object value now being referred to as a 'literal'. Literals cannot be searched in the semantic search function, but they can be found in the plain text search.

Additional Topics:

Regions:

You can create either rectangular shapes or freeform geometric shapes by clicking one of the two region buttons below the image in the editor. Click the point where you want to begin the region, draw across the region, and click again to complete your shape.

Creating more complex triple-tags or deleting a triple-tag:

Hover at the end of the triple-tag until you see the plus and minus signs and select the minus sign to delete the triple.

The plus sign lets you add a 'which' clause to chain together triple-tags and build more complex expressions. Please see the section on Complex Expressions: the which clause for more information on this topic.

Batch Editing Multiple Images:

Choose the "Multi-select" button and then click on each image you want to include in this selection. Selected images will have a blue background.

Click on an image again if you want to de-select it from the group.

When you have made all of your selections, click on the "Edit Selected Images" Button. Thumbnails of the selected images will appear at left in a new working window. Any triple-tags, details, keywords or descriptions entered here will be saved with all of these images. Select "Save and Close" when you are finished (Note: this function works, but we are currently debugging a faulty error message that appears during this process).