2.1.2 Annotator

This topic describes the information about the annotator.

Annotation is the process of identifying information within a documented content and tagging them as a specific type of information. Each use case defined, have their own relevant maintained list of tags/entities, which is used to annotate source documents for a use case.

Figure 2-3 Loan Draw Down

Description of "Figure 2-3 Loan Draw Down"

• Annotator
  This topic describes the systematic instructions to perform the annotations on a source document for a use case.

2.1.2.1 Annotator

This topic describes the systematic instructions to perform the annotations on a source document for a use case.

Specify User ID and Password, and login to Home screen.

1. On Home screen, click Machine Learning. Under Machine Learning, click NLP Toolkit.
2. Under NLP Toolkit, click Annotator.
   The Annotator screen displays.

   Figure 2-4 Annotator

   
   

   
   Description of "Figure 2-4 Annotator"

   
   
3. Specify the fields on Annotator screen.

   Note:

   The fields, which are marked with an asterisk, are mandatory.

   For more information on fields, refer to the field description table.

   Table 2-3 Annotator – Field Description

   Field	Description
   Action Type	Select require action type. The available options are: Create New Annotated File Edit Created Annotated File
   Source File Definition	Select the source document from local windows explorer based on the Action Type selected.
   Document Type	Displays the list of all the use cases defined under use case definition.
   Get Labels	Displays the maintained Tags/entities for the selected Document Type.
   Create Annotated File	Once annotations of all the Tags are completed, this performs two outcomes as below, Create annotated text file in the defined NER train path as maintained under use case definition. Create text file in the defined DOC train path as maintained under use case definition.

Annotate the Source Files

4. Select Create New Annotated File in Action Type.
5. Click Select File. It will open the windows explorer. Navigate and select the source document to be annotated.
6. The source document displays in the Original File field and text version displays in the Text Form field.

   Figure 2-5 Annotator - Text Form

   
   

   
   Description of "Figure 2-5 Annotator - Text Form"

   
   
7. Select the Document Type from drop-down list.

   Figure 2-6 Annotator - Document Type

   
   

   
   Description of "Figure 2-6 Annotator - Document Type"

   
   
8. Click Get Labels.
   It loads all the maintained tags for the Document Type.

   Figure 2-7 Annotator - Tags

   
   

   
   Description of "Figure 2-7 Annotator - Tags"

   
   
9. Identify and select information within the Text Form section of the document.
10. Right click to display the list of tags and select the relevant tag.

    Figure 2-8 Annotator - List of Tags

    
    

    
    Description of "Figure 2-8 Annotator - List of Tags"

    
    

    Figure 2-9 Annotator - Select Annotation Label

    
    

    
    Description of "Figure 2-9 Annotator - Select Annotation Label"

    
    
    The selected tag and the information appears in section Annotations under Tag Name and Tag Value.

    Figure 2-10 Annotator - Annotations

    
    

    
    Description of "Figure 2-10 Annotator - Annotations"

    
    
11. Repeat the above steps for all the displayed tags as per availability of information in the source document.
12. Select a Tag Name from the Annotations section and RIGHT- CLICK to delete the Tag Value.

    Figure 2-11 Annotator - Tag Value

    
    

    
    Description of "Figure 2-11 Annotator - Tag Value"

    
    
13. Once all the tags are assigned the relevant information, click Create Annotated File to create the annotated file and end the process.