Text Analytics

Modified on Thu, 01 Feb 2024 at 08:39 AM

Purpose

Review terms in cases that highlight behavioral trends or plot customer sentiment scores against product categories release versions, support regions, time zones, case data, etc., to gain perspective on why case volume or resolution time is impacted.  Identify keywords and phrases in your cases and correlate them against other trends that might lead to new insights.



Summary

Leverage highly interactive and configurable charts to track patterns, mark events (shown below as a red flag) to note unusual spikes in volume, increased negative sentiment, and more. You can export an image of the main chart and the underlying data as a CSV. 



Text Analytics

The main chart presents a high-level summary of case trends over the last week, month or quarter.  

Under the main chart, are tabs for the Overview of cases, Negative Sentiments, Escalation Requests, Feature Requests, Escalations, and other metrics. Configuration of what tabs are shown versus not shown is defined under Control Center > PAGES > Text Analytics.


Events

If you wanted to note an uptick in case volume to add an event:


  1. Click once on the chart
  2. Pause, then click again
  3. Enter your comments and and select a flag color
  4. Press enter


Changing the grouping of the events does not impact the event you have marked. The example to the right just happens to be grouped by Operating System.





Tracking Trends

Text Analytics can reveal insights that otherwise would not be uncovered with traditional reporting tools. For example, your organization has the following product lines: Fedora Linux, Mac OS, Microsoft Windows, RHEL and Ubuntu Linux and you believe one of these operating systems is causing far more frustration with your users than others. To test this assumption, you do the following:


  1. Group the chart by Product
  2. Add keywords "annoyed" and "frustrated" and hit ENTER
  3. It may be difficult to get a clear picture of which Product have more of the terms "annoyed" and "frustrated" in cases based on the ebb and flow of the chart


  1. To delve further into the review, you select Negative Sentiments (beneath the chart)
  2. Update the Showing All Signals to only Inbound Signals (so that you are only viewing terms that were voiced by customers)
  3. Then group the results by Product


This quickly reveals that Ubuntu Linux had 95 cases where "frustrated" and "annoyed" appeared whereas most other products had a much lower number of cases where those key words appeared within case text.






Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select atleast one of the reasons

Feedback sent

We appreciate your effort and will try to fix the article