Smart Clustering Themes Dashboard

Updated 

Tracked Vs Untracked Themes  

The Smart Clustering Themes feature includes two separate dashboards: Voice of Customer and Voice of Advisor. These dashboards are divided into two main sections: 

  1. Tracked Themes: This section contains keywords and phrases from conversations where the customer's intention is identified. 

  2. Untracked Themes: This section includes keywords and phrases from conversations where the customer's intention is not identified. 

Dashboard Filters 

Keyword Search – Allows users to search of a keyword or phrase from the clusters. 

From User Social Network – Filter clusters basis the social network 

Account - Filter clusters basis accounts  

Sentiment - Filter clusters basis the sentiments – neutral, negative, positive 

Smart Clustering Keyword – Filter using the clusters/themes 

Smart Clustering Keyword Type  

  1. Subject – Refers to the main theme of the cluster. For example: Credit Card is a smart clustering keyword type 

  2. Attribute – Refers to the sub-themes/intent associated with the subject.  

Hide/Archive Smart Clusters 

Users can hover on the smart cluster, where a distribution of the phrase across sentiment is shown. Click on “Hide” button to archive any smart cluster.  

Unhide/Unarchive Smart Clusters 

Click on “Show All” in the VoC/VoA widget and select the Hidden filter as “True”. All the hidden/archived smart clusters is visible here.

Co-occuring Themes

Users can use the drill-down feature to have a deeper analysis of the clusters. Double-click on the clusters to view the sub-clusters/themes related.

Details on any Cluster

Click on the cluster once to open third pane which provides detailed insights on case associated with the cluster, sentiment distribution and intents detected in those cases

  1. Case Volume Trend – Provides volume trend of cases that are associated with the cluster/sub-cluster.

  2. Case Sentiment Analysis - Provides sentiment wise distribution of cases along with smart clustering case count that are associated with the cluster.

  3. Break-down by Source - Provides source wise distribution of cases that are associated with the cluster. 

  4. Case Details – Provides details of all the cases that are associated with the cluster/sub-cluster.