New filtering options for AI Text Analytics

We are excited to introduce the next set of improvements to AI Text Analytics, making it easier to explore, refine and work with insights from Open Text responses.

You can now create filters based on AI Text Analytics Topics and Sentiment. This makes it possible to filter report results by selected Topics or Sentiment categories, helping users move from an overall view of the feedback to the parts that are most relevant to their analysis. Filters can be used to explore specific Topics, compare Sentiment categories and examine different combinations of Topics and Sentiment across the responses.

The filter setup has been improved with new sorting options, making it easier to review and manage the available values. Users can sort values alphabetically or based on whether they are selected, which is especially useful when opening and editing an existing filter. This provides a clearer overview of what the filter contains and makes it quicker to adjust the selected Topics or Sentiment values.

Together with the recently released improvements, this completes the current phase of AI Text Analytics development. AI Text Analytics Topics and Sentiment can now be included as additional columns in raw data reports. In Livereports, AI Text Analytics items can be filtered using value and interval filters based on other survey data, such as selected answer alternatives. AI Text Analytics items can also be placed in the preferred order in Report Builder. With the new ability to create filters directly from Topics and Sentiment, AI Text Analytics results are now easier to export, organize, filter and use throughout the reporting experience.

Learn how to create filters based on AI Text Analytics Topics and Sentiment in our Academy.

We have also included a number of bug fixes and improvements in this release:

• Respondents no longer receive a ”quota full” message after responses have been deleted and the quota has available capacity again.
• Fixed an issue where a respondent changing an option in a Hierarchy question could cause validation to hang and prevent them from continuing.
• Alternatives used in filter conditions are now shown in the Quest default language instead of the service language.
• The connection between the Essentials Quest ID and the New External API Quest ID (GUID), used when working with the API, is now visible.
• Fixed an issue where the timeline in NPS exports could be incorrect when filters or comparisons were applied.
• Email addresses can now be removed from the opt-out list even when the list contains more than 100,000 addresses.
• The Quests-QuestionsList API call no longer returns a server error when a Quest contains an NPS Slider question.
• Fixed an issue where slider callouts in matrix questions could be cut off or misaligned.
• Removing text from the Quest Introduction no longer leaves HTML behind that can cause the system to treat the introduction as if it still contains content.
• Topic language is now displayed as the language only.

We hope you enjoy the improvements and find them useful when working with AI Text Analytics. As always, thank you for using Questback!