---
sourceDocument: Xanadu Customer Service Management
sourceDocumentLink: https://servicenow-prod.fluidtopics.net/r/xanadu/customer-service-management

 Release :

    - xanadu

ft:locale :

    - en-US

ft:publication_title :

    - Xanadu Customer Service Management

ft:clusterId :

    - csm

bundleId :

    - csm

workflow :

    - Customer and Industry


---

# Create a model to predict case sentiment

# Create a model to predict case sentiment {#ariaid-title1}

* Release version: Xanadu
* 
* Updated August 1, 2024
* 
* ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 1 minute to read

Edit and test the pre-trained sentiment model to predict sentiment for customer service cases.

Before you begin
:   Role required: ml_admin, admin

About this task
:   The case sentiment model is pre-trained with a large data set to learn communication patterns. You select a set of records to test the model and then view the results before deploying.

## Set up your model {#csm-task-intel-create-sentiment-solution__section_icx_xdz_g5b}

1. Navigate to AllTask Intelligence for Customer ServiceSetup to access the Task Intelligence Admin Console.
2. Select Edit model on this model: Predict case sentiment to improve CSAT.

   This opens the model and displays the first of five pages. Each page in the model asks you questions
   and helps you select the information you need to build an effective model.
{#csm-task-intel-create-sentiment-solution__ol_lzt_g2z_g5b}

## Define the purpose {#csm-task-intel-create-sentiment-solution__section_a2s_1lv_h5b}

Tell the model when you want it to make predictions.

The sentiment model is pre-trained with a large data set to understand communication patterns. This data comes from customer emails and from case descriptions and comments.

1. Select the table that contains the records to be tested.Select All cases to test records in the Case table or select a case type table.

2. Select Save \& continue.
{#csm-task-intel-create-sentiment-solution__ol_r4b_klv_h5b}

## Test your model {#csm-task-intel-create-sentiment-solution__section_bvg_mlv_h5b}

Choose the cases to use to test the model.

Selecting this information tells the model what type of records to use for testing. When the model is deployed, it predicts sentiment on all records in the table that match the selected conditions.

1. Use conditions to choose a set of records for testing.
2. Select Launch testing.
{#csm-task-intel-create-sentiment-solution__ol_yd2_rlv_h5b}

## Assess your model {#csm-task-intel-create-sentiment-solution__section_pxw_tlv_h5b}

Assess the results from the testing and view sample results for past cases.

Reviewing the results gives you a preview of how your model will perform after being deployed.

1. Select View all test results to see a list of results.
2. Select Save \& continue.
{#csm-task-intel-create-sentiment-solution__ol_vxd_1mv_h5b}

## Set your preferences {#csm-task-intel-create-sentiment-solution__section_nkz_bmv_h5b}

Tell the model how you want to use the sentiment predictions.

1. Display predictions in the Current sentiment and Sentiment over time fields.
2. If desired, you can run the model in the background only.With this option, the system makes the predictions and stores the information in the Predictions History but does not add any information to the case
   records.

3. Select Save \& continue.
{#csm-task-intel-create-sentiment-solution__ol_n34_2mv_h5b}

## Deploy your model {#csm-task-intel-create-sentiment-solution__section_ims_gmv_h5b}

Review your choices from the previous pages. Then you can select Deploy to deploy the model.

*[\>]: and then