Text Classification and Entity Extraction are two more dynamic forms that have been introduced in step 2 of the project setup sub-menu,where one configures queues. Other enhancements are also introduced. Explanation follows.

Figure 1

Configuration-->Project Setup-->Step 2 Queues

Activate /Deactivate is the new header to compliment the new features that have been incorporated.

The Form settings page and form type have been realigned.

Figure 2

Entity Extraction

The latest release comes with a very important feature,namely,Entity extraction .It is listed as a choice in the form type's drop-down.

Using this option,one can label names of people,organizations,geographical locations,etc. First by feeding the input text in the system,creating labels(Fields) and training the system to categorize the text into the created labels.Thus, this release presents NER (Named Entity Recognition)capabilities.After a period of time ,labelling important elements related to projects can be trained and over time, system itself will label it as configured.

Select  Entity extraction from the Form Type.Select yes to send the text to My Worklist queue and to enable the different form settings. BackPush  option can also be availed.

Click Here in the Basic Form Setting will facilitate the following form configuration process.The selected queue for entity extraction is process and accordingly the Process' form configuration will be enabled.Refer Figure 4.

Figure 3

Check the appropriate boxes. In the following scenario,a paragraph comprising several sentences are fed as input text in the processing queue called Process. The configured fields will serve as labels and can be used to classify a single or multiple words.Click the OK button to submit. 

Figure 4

 The input text will be auto populated in the incoming queue.In the following example,next queue as per configuration is Process. Submit.

Figure 5

It will go to the Process queue.

Go to My Worklist sub-menu.

Click on the required URN.

Figure 6

Classify single or multiple words by selecting them  and clicking on the respective fields or labels that is seen on the right side of the screen in  different colors.Submit.

Figure 7

Text Classification

The same steps are to be followed for text classification.Only instead of selecting words,entire paragraph of text is selected and classified using the configured fields or labels.

Make the appropriate selections and select Text Classification in the Form Type.Note that the completed queue is selected for this process in the following example.

Figure 8

Click Here to enable the Completed queue's Form Configuration page.Check the input box against the pertaining field name.Check the other fields that need to act as labels for classifying paragraphs.Click on the OK button.

Figure 9

 Go to the Incoming queue.The text will be auto populated.Submit and go to My WorkList to select the required URN.

Proceed to use the configured fields to classify the input content as paragraphs.

Figure 10

Note: Input words,text or content is auto populated. This is achieved in the back end, using OCR capabilities.

Incoming queue is exempt for both Entity extraction form and Text classification forms.Only Basic forms can be used for the index or incoming queue.

Choose max to allow for unlimited text to be given as input text.

Double click on text or words  and use the delete key to delete the same.