LinkedIn
Copied!

Table of Contents

Managing models in text predictions

Add topic models and entity extraction models to your text prediction, and define whether to run the model on the whole text or only parts of it. Review and download information about your text models, such as the model binary file, taxonomy and training data, performance reports, and feedback.

  1. Open the text prediction:

    1. In the navigation pane of App Studio, click Channels.

    2. In the Current channel interfaces section, click the icon that represents a channel for which you want to configure the text prediction.

    3. On the channel configuration page, click the Behavior tab, and then click Open text prediction.

  2. In the Prediction workspace, click the Models tab.

  3. To modify the scope of the text on which you want to run the model, in the Data column, click the Edit icon, and then select one of the following options:

    • Body
    • Attachment
    • All
    Use this setting with email channels to assign different models in the text prediction to run on the body of the email, on the attachments, or both.
  4. To add an existing topic model or entity extraction model to the prediction:

    1. In the top-left corner of the Topic or Entity section, click Add.

    2. In the Add model window, select Pick an existing model from a list, and then click Next.

    3. Select a model from the list, and then click Add.

  5. To create a new topic model or entity extraction model and add it to the prediction:

    1. In the top-left corner of the Topic or Entity section, click Add.

    2. In the Add model window, select Create a new model, and then click Next.

    3. Create the model:

    4. To remove a model from a prediction, click the Delete icon to the right of the model, and then confirm by clicking Submit.

  6. To review the model data, click Model data.

    The Model data tab displays the information about the existing test data, existing and pending training data, and the f-score for each model.
  7. To download the data for a model, click Download to the right of the model name.

  8. Click Save.

    Related Content

    Have a question? Get answers now.

    Visit the Collaboration Center to ask questions, engage in discussions, share ideas, and help others.