In the Configure visualization side panel and select Configure to control which concepts appear, how match percentages are calculated and how concept volume is displayed. The controls may appear as condensed dropdowns in the current interface.
Choose which concepts appear, how matches are calculated, and how results are displayed. Configure, formerly called Configure Visualization, controls the content and presentation of the current analysis view. It is available in Volume, Galaxy, Drivers and Sentiment.
Use Configure to adapt each feature to the question you want to answer rather than relying only on its default view.
Configure generally helps you decide:
Which concepts should appear?
Which records should be included in the calculations?
How should the results be displayed?
The available options vary by feature and by the metadata included.
Choose which concepts to visualize
The first setting determines which part of the text analysis project you want to examine.
Depending on the feature and available data, you may see the following options.
Top concepts by prevalence
Display concepts that are especially characteristic of the text analysis project compared with language in general.
Use this option when:
Exploring a text analysis project
Exploring without a predefined topic
Looking for language that distinguishes the text analysis project
Identifying concepts to add to a concept list
Prevalence is different from total match count. A concept can be distinctive to the text analysis project without being the most frequently mentioned.
Conversation clusters
Display concepts from the automatically detected clusters of conversation.
Use this option when you want to explore the major themes. It is especially useful in Galaxy, where you can see how the concepts within and between clusters relate.
Concepts linked to scores
Display concepts associated with differences in a selected numeric field.
Choose the relevant score field and the number of concepts you want to display.
Use this option when the text analysis project contains a meaningful outcome such as a rating, satisfaction score, NPS, or another numeric measure. Continue the analysis in Drivers before interpreting these concepts as important influences.
Concepts linked to strong feeling
Display concepts associated with especially positive or negative sentiment.
This is a useful starting point in Sentiment, but sentiment strength should be interpreted alongside the number of matching records. A concept may show strong sentiment while representing only a small portion of the text analysis project.
Unique to this filter
Display concepts that are more characteristic of the currently filtered records than of the rest of the text analysis project. You must apply a filter before this option becomes available. Use it to explore what distinguishes:
Low- and high-scoring responses
Products or brands
Regions or customer groups
Departments or employee groups
Channels
Time periods
A concept identified as unique is not necessarily exclusive to the filtered group. It is more distinctive of that subset than of the remaining data. If no concepts are returned, the group may not contain language that is meaningfully different from the rest of the text analysis project. An empty result can therefore be informative.
Active concepts
Display concepts you or your team have already saved and organized. Use this option when you want to:
Monitor an established set of topics
Apply a business taxonomy
Compare the same concepts across features
Reuse AI-generated Themes
Maintain consistent colors in an analysis or report
Select the applicable concept list from Concepts before using Active concepts.
Choose how match percentages are calculated
When filters are applied, Configure lets you choose the denominator used for match percentages.
Based on currently set filters
Calculate the percentage using only the records included in the active filters. Use this when asking: What percentage of this selected group discusses the concept?
For example, if you filter to one product, the percentage represents the share of that product’s records that match the concept.
Based on total records
Calculate the percentage using all records in the text analysis project, including records outside the active filters. Use this when asking: How much does this selected group contribute to the text analysis project’s total conversation?
The same match count can produce different percentages depending on the denominator. Always confirm this setting before comparing or reporting percentages.
Include or exclude conceptual matches
When available, select Include conceptual matches to count both direct mentions and closely related meanings. Keep conceptual matches included when you want to understand the broader conversation around a topic.
Clear the option when your question specifically requires explicit mentions of the concept. Excluding conceptual matches creates a narrower, search-like result, but it may omit relevant records that use different wording. If a result appears unexpectedly broad, review its matching records before excluding conceptual matches entirely.
Configure Volume
Configure Volume
Volume includes settings for how concept reach should be displayed.
Show prevalence rank
Display each concept’s relative prevalence within the text analysis project.
Keep this visible when you want to compare how distinctive a concept is with how many records mention it. Hide it when the analysis is focused only on counts or percentages.
Choose the Volume display
Volume can be displayed as:
Overall concept volume for a direct comparison of counts and percentages
Over time for tracking changes using a date field
By category for comparing products, regions, channels, groups, or other categorical metadata
By score value for comparing individual numeric scores
Across a numeric range for grouping fields with many numeric values
Choose the display based on the question. For example, use Over time to identify an emerging issue and By category to determine where that issue is concentrated.
Choose bar chart colors
Volume can use:
Default colors to distinguish exact and conceptual matches
Concept colors from the selected concepts or concept list
Sentiment colors to add emotional context
Changing the colors does not change which records match the concepts.
Configure Galaxy
Configure Galaxy
Galaxy includes options for controlling the density and visual organization of its relationship map.
Choose the number of concepts
Use fewer concepts for a clean, high-level view or presentation. Display more concepts when you need to:
Explore an unfamiliar dataset
Find smaller clusters
Discover less prominent language
Build or refine a concept list
A more detailed Galaxy can reveal additional relationships, but it may also be harder to interpret. Begin with a focused view and increase the number of concepts when additional detail supports the analysis.
Choose how concepts are colored
Galaxy can color concepts by:
Conversation clusters to distinguish automatically detected themes
Active concepts to emphasize concepts currently being tracked
Concept lists to apply an existing analytical framework
Use conversation-cluster colors for discovery. Use concept-list colors when presenting a curated business taxonomy or comparing established themes.
Color organizes the visual narrative; it does not change the strength of the underlying relationships.
Configure Drivers
Configure Drivers
Drivers includes settings for selecting and comparing numeric outcomes.
Choose the axis score field
Select the numeric field that should appear on the Drivers vertical axis. The field should represent a meaningful outcome, such as:
Rating
Customer satisfaction
NPS
Employee engagement
Purchase intent
Time to resolution
Another relevant numeric measure
Synchronize the score fields
Keep Sync concepts score field and axis score field selected when the concepts and vertical axis should use the same numeric outcome. Clear it only when you intentionally want to identify concepts linked to one score and examine them against another.
For example, you might select concepts associated with a product’s Ease of Use score and plot them against Overall Rating. Clearly document both fields so the result is not misinterpreted.
Compare periods
When available, enable Compare period over period and select a date field to examine whether concept associations have changed.
Use comparable periods and check the number of records in each. A difference may reflect seasonality, collection volume, or a different population rather than a change in the experience itself.
Configure Sentiment
Configure Sentiment
Sentiment uses the shared concept-selection and percentage settings to control which topics appear and which population is represented. For broad discovery, start with Concepts linked to strong feeling. For recurring analysis, use Active concepts to compare the sentiment of an established set of themes.
Sentiment does not use a numeric score axis. It analyzes the positive, negative and neutral language surrounding each concept. Review sentiment percentages alongside the number of matching records, then select a concept to examine the supporting examples. Use Download more matches when you need additional examples for the selected concept.
Combine Configure with Filters
Configure and Filters answer different parts of the analytical question:
Filters determine which records are included.
Configure determines which concepts appear and how the resulting view is calculated and displayed.
For example, to discover what distinguishes low-rating feedback:
Open Filters
Select the applicable low-rating values
Return to Configure
Select Unique to this filter
Review the resulting concepts
Open Records to validate the language behind them
Repeat with the high-rating group for comparison
Change one filter or configuration setting at a time so you can understand what caused the visualization to change.