Configure Datasets

Required permission: Interaction Analytics Edit

Create a Dataset

You can apply a category template, a workspace template, or both when you create a dataset. If you plan to apply new templates, you must create them before you create the dataset. Datasets can include any supported interaction type:

  • Inbound and outbound voice
  • Omnichannel and Digital First Omnichannel chat
  • Omnichannel and Digital First Omnichannel email
  • CXone Messaging SMS
  • Digital First Omnichannel channels:
    • Apple Business Chat
    • Facebook Messenger
    • Google Business Manager
    • Line Messaging
    • Microsoft Teams
    • Slack
    • Telegram Messaging
    • Twitter Direct Messaging
    • Viber
    • WhatsApp
    • Custom channels using the Bring Your Own Channel feature
Image of the create dataset page
  1. Click the app selector and select Analytics.

  2. Click My Data > Analyze New Data. Click the options icon options icon and then click Edit.

  3. Click the Channels drop-down and select all channels you want to include in the dataset. The preceding image selects multiple channels.

  4. Give the dataset a descriptive Dataset Name. Specify the Dataset Period you want the dataset to include and apply the category template and workspace template you want.

  5. If you want to filter the dataset data to only include contacts that include or exclude certain keywords, phrasesClosed Combination of words that have special significance when used together in a specified order, such as "want to cancel"., entitiesClosed Defined keyword or phrase, with or without variants, in an interaction or message, sentimentsClosed Overall mood or result of the interaction as determined by analysis of words, phrases, and context of the transcript. Interactions can be positive (blue), negative (red), frustrated, mixed (dark gray), or neutral (light gray)., metrics, or a combination of these things, use the All of These, At Least One of These, and None of These fields as needed:

    • Click and drag sentiment, metric, frustrated, and unresolved tags from Add criteria to the preferred fields and configure them as needed.

      Tag Description
      Sentiment

      Pulls data into the dataset based on the sentiment assigned to the interactions.

      1. After you drag a sentiment tag to the preferred field, click the agent/contact icon on the left side of the tag (indicated by a headset or person icon, or both) to toggle the options between agent sentiment, contact sentiment, or agent and contact sentiment.
      2. Click the down arrow on the right side of the tag to open a configuration pop-up.
      3. Select the Type of sentiment you want to use as criteria for your dataset between End Sentiment at the conclusion of the interaction or the Overall sentiment throughout the interaction.
      4. Select the preferred Sentiment from Positive, Negative, Neutral, and Mixed.
      5. Click Done.
      Metric

      Pulls data into the dataset based on whether it meets a certain metric value or threshold.

      1. After you drag a metric tag to the preferred field, configure the metric criteria in the pop-up. If a pop-up did not appear when you dropped the tag or you closed it, click the down arrow in the tag to open it again.
      2. Set the Field to the metric you want use as the criteria. See the Metric Widget for details about each available metric.
      3. If you selected a metric that produces a numeric value, select an operator from the drop-down under Value from Equals, Greater than, Less than, Greater than or equal to, or Less than or equal to and enter the number threshold you want the operator to compare with the metric in the text field. If you selected a metric that produces a string value, enter the value you want the metric to match in the Value text box. If you selected a metric that produces a boolean value, select True or False from the drop-down.
      4. Click Done.
      Frustrated Pulls data into the dataset based on whether the contact became frustrated during the interaction.
      Unresolved

      Pulls data into the dataset based on the resolution status (Unresolved or Resolved) at the end of a contact.

      1. After you drag an unresolved tag to the preferred field, click the down arrow on the right side of the tag.
      2. Specify in the Type field whether you want Resolved or Unresolved contacts to be the criteria.
      3. Click Done.

    • Enter text directly into the Find contacts with fields to enter keywords, phrases, or entities.

  6. For each tag you enter, you can specify the following:

    • Whether the keyword, phrase, or entity should appear in the agent or contact side of the interaction, or both, by clicking the agent or contact icon on the left side of the tag and selecting the criteria you want to appear.
    • Whether the keyword, phrase, or entity should be an exact match to what you entered or a variation of what you entered, by clicking the proximity drop-down on the right. An exact match only returns the term or phrase exactly as you entered it. If you choose to match with variations it would also return variations of that word, for example if you enter activate, it could return interactions that include activate, activates, activating, activated, and so on.

    When you enter a phrase, clicking the down arrow also allows you to specify how close the words in the phrase must be to each other for the contact to be considered a match on the phrase. The default is Exact Phrase, meaning the words must be together and in the right order. You can also specify that the words must be one or fewer words apart to eight or fewer words apart and not necessarily consecutive. For example, if you enter the phrase "rude agent" and select a proximity of 3 Words, a transcript that says "an agent was disrespectful and rude" would count.

  7. Click Save.

View Dataset Categories

To better understand how contacts are placed into categoriesClosed Data groups that Interaction Analytics uses for parsed call transcript data. They make it easier to find trends., you can view the category tree and the rulesClosed Criteria used to determine whether a transcript is placed in a certain category. Rules contain sets of keywords or phrases a transcript must include or exclude. associated with each category.

Specifies which media types you want to include in the dataset for analysis. You can include any combination of Voice, Email, and Chat contacts. Selecting at least one channelClosed A way for contacts and agents to interact, such as voice, email, chat, social media, and so on. is required to save the dataset.

Image of dataset categories.

  1. Click the app selector and select Analytics.

  2. Click My Data, then click the name of the dataset you want to open.
  3. Click Options > Edit Categories. Drill down into a category of interest by clicking the category name. You can open and click subcategories to drill down as deep as you want. To quickly expand or collapse all categories, click the arrow in the top left corner of the category tree workspace.
  4. If no contacts appear, ask an administrator to load a contact sampling for that dataset.

  5. To view the criteria required for a contact to fall into a certain category, click the Rules tab.

    A list of rules appears to define the contacts of this category. In the preceding image, a contact must have negative sentiment and the word "package" or have negative sentiment and the word "shipping" to fall under the Negative Shipping category. Some of the pre-built rules are hidden so you cannot dig in to see exactly what it is doing.