ML Predictive Models & ML Text Classification

Machine learning (ML)

Machine learning is an application of artificial intelligence (AI) that focuses on the development of computer programs that can access data and use it to learn for themselves. Comidor provides ML models such as Predictive ML Models and ML Text Classification Models in order to enhance Digital Process Automation. Use existing data to train your ML Models and get predictions on specified user fields. With Comidor ML Text classification model, assign tags or categories to text according to the field content.

 

With Comidor ML models you can:

• Enhance process automation
• Eliminate errors
• Save processing time

Comidor ML Text classification can be used for:

• Topic labeling
• Spam detection
• Intent detection

ML Predictive Models

To access ML Predictive Models, go to App Factory Icon > RPA & AI/ML > Predictive Models.

1. Click on the + Icon at the top of the screen to open the Create Form.
2. Type a Title for your ML Predictive Model.
3. Link the Model with a Connected Application or select the respective Entity.
4. Choose your preferred Classifier: 
   • Comidor offers a variety of algorithms that sort data into labelled classes or categories of information, such as J48, DecisionTable, DecisionStump, etc.
5. The Target field refers to the field you want to make the prediction for. Depending on the entity you have selected above, it will populate relevant fields to choose from.
6. Add 2 or more Training fields:
   • The values of these fields will be used by the classifier.
   • The user fields available here are linked to the entity you have chosen.
7. Limit training data with a condition to define specific values to be used by the classifier.
8. Select the desired Save option (refer to the Quick Reference Guide).

Edit ML Predictive Models

1. Go to App Factory Icon > RPA & AI/ML > Predictive Models.
2. Select the ML Predictive Model to edit.
3. Click on the Edit button to open the Edit Form.
4. Edit the information you want and click on the desired Save option (refer to Quick Reference Guide)
   

   

Train ML Predictive Models

After creating your ML Predictive Model, you have to train the Classifier based on the data entered in Comidor. The training process might take a while. The more data we have, the better the Accuracy we achieve.

1. Go to App Factory Icon > RPA & AI/ML > Predictive Models.
2. Select the ML Predictive Model to train.
3. Click on the Train button. Click on this option as many times as you wish, until you reach an acceptable Accuracy rate.
4. Click on the Save button that is appearing next to the Accuracy percentage. The ML Predictive Model is ready to be used in a workflow.

Delete ML Predictive Models

1. Go to App Factory Icon > RPA & AI/ML > Predictive Models.
2. Select one or more ML Predictive Models records.
3. Click on Delete to delete one or multiple ML Predictive Models at the same time. A confirmation pop-up box appears.

 

ML Text Classification

To access ML Text Classification, go to App Factory Icon > RPA & AI/ML > Text Classification.

1. Click on the + Icon at the top of the screen to open the Create Form.
2. Type a Title for your ML Text Classification.
3. Link the Classification with a Connected Application or select the respective Entity.
4. The Target field refers to the field you want to make the prediction for. Depending on the entity you have selected above, it will populate relevant text fields to choose from.
5. Add Training field:
   • The values of this text field will be used by the Text classifier.
   • The text user fields available here are linked to the entity you have chosen.
6. Limit training data with a condition to define specific values to be used by the Text Classifier.
7. Select the desired Save option (refer to the Quick Reference Guide).

Edit ML Text Classification

1. Go to App Factory Icon > RPA & AI/ML > Text Classification.
2. Select an ML Text Classification to edit.
3. Click on the Edit button to open the Edit Form.
4. Edit the information you want and click on the desired Save option (refer to Quick Reference Guide).

Train ML Text Classification Model

After creating your ML Text Classification model, you have to train the Classifier based on the data entered in Comidor. The training process might take a long time to fetch accurate results. The more data we have, the better Accuracy we achieve.

1. Go to App Factory Icon > RPA & AI/ML > Text Classification.
2. Select the ML Text Classification Model to train.
3. Click on the Train button. Once you click on this option, you can see the Accuracy rate of the Text Classifier. Click on this button as many times as you wish, until you reach an acceptable Accuracy rate.
4. Click on the Save button. The ML Text Classification Model is ready to be used in a workflow.

Delete ML Text Classification

1. Go to App Factory Icon > RPA & AI/ML > Text Classification.
2. Select one or more ML Text Classification records.
3. Click on Delete to delete one or multiple ML Text Classifications at the same time. A confirmation pop-up box appears.

---

 

Add a Predictive ML in the workflow design, to trigger an ML Predictive Model and get a prediction for the Target Field of a workflow, or a Text Classification to classify the selected Target field based on the Text Classification Model.

To access Workflows go to App Factory Icon > Workflow Automation > Workflows

Predictive ML

• Drag-and-drop the Predictive ML.
• Give a Title to the component.
• Give the Parent Stage which is the stage of the parent process as soon as this step is reached.
• Select which Model you would like to run at this step, from the list of the ML Predictive Models that you have already created.

Text Classification

• Drag-and-drop the Text Classification element.
• Give a Title to the component.
• Give the Parent Stage which is the stage of the parent process as soon as this step is reached.
• Select which Model you would like to run at this step, from the list of the ML Text Classification Models that you have already created.
• You will also see the Target field and the Training field(s) that were selected in the ML Text Classification Model in view-only mode.

 

---

Find out more on how to create and manage workflows step by step and learn all about Comidor Workflow Elements.