Expert.ai - How to get started with Expert.ai in Automations

[Ionut_Dumitrascu]

This article gives an overview of the available blocks in the Expert.ai connector in Qlik Application Automation. It will also go over a basic use-case which involves applying a Sentiment analysis over a given text using this connector.

This connector consists of 4 blocks:
- Run Analysis Task
- Run Classification Task
- Run Geotax Categorization Task
- Run Information Detection Task

The Run Analysis Task block can be used to conduct the following types of text analysis:

• Deep linguistic analysis - resolve ambiguities in the text by identifying the relationships between concepts and assigning a specific meaning to each word
• Named entities recognition - understanding the most important entities to extract
• Relation extraction - to identify the relationships between the subject, action, and object words
• Sentiment analysis - to understand if the tone is positive or negative
• Full document analysis - includes all of the above

The Run Classification Task block can be used to:

• Categorize an article based on embedded taxonomies including IPTC Media Topics and a geographic taxonomy
• Leverage the emotional and behavioral traits taxonomies to identify emotions displayed in text and personality traits associated with a person

The Run Geotax Categorization Task block can be used to get a detailed 'geotax' categorization for a given text.

The Run Information Detection Task block can be used to:

• Automatically identify and classify personally identifiable information (PII) that needs to be protected to maintain compliance with GDPR and other privacy regulations
• Provide information on the readability of a text, vocabulary richness, document structure, and grammar

Authentication for this connector is basic, using username and password as parameters.

Let's now go over a basic example of how to use the Expert.ai connector:

1. Create an automation;
2. From the left menu, use the Search for blocks input to search for the Inputs block;
   
3. Drag and drop the block into the automation and link it to the Start block;
4. On the right side of the screen, click on the Label and add a name for the input field. Choose Multiple input lines as type.
   
5. From the left menu, select the Expert.ai connector;
6. Use the Search for blocks input to search for the Run Classification Task block;
7. Drag and drop the block into the automation and link it to the Inputs block;
   
8. On the right side of the screen, click on the Taxonomy input field, click on the do lookup option, then choose emotional traits. Click on the Language input field, and choose your preferred language. Click on the Document Text input field and select your input;
   
9. From the left menu, search for the Loop block;
10. Drag and drop the Loop block into the automation and link it to the Run Classification Task block;
11. On the right side of the screen, click on the Loop over items of list input field, click on the output from Run Classification Task, and then click on the Categories list;
    
12. From the left menu, search for the Output block;
13. Drag and drop the Output block into the automation and add it inside the loop.
    
14. On the right side of the screen, click on the input field, click on the item in Loop option. Do a Test run and then click on the label.
    
15. Run the automation. Add your text to be evaluated in the input field. This will output a list of emotions that the Expert.ai analyzer found in your text.

This example is added as an attachment.

The information in this article is provided as-is and to be used at own discretion. Depending on tool(s) used, customization(s), and/or other factors ongoing support on the solution below may not be provided by Qlik Support.