Open the side panel and select Configure visualization to adjust the Galaxy map. In the current interface, several settings appear in condensed dropdowns. Configuration changes which concepts and visual organization you see. It does not change the underlying relationships learned from the text analysis project.
Choose the right level of detail
Adjust the number of displayed concepts according to your purpose.
Use a smaller number for:
A high-level introduction
A cleaner presentation
A focused discussion with stakeholders
Exploring the strongest relationships around one topic
Use a larger number for:
Detailed discovery
Finding smaller conversation clusters
Exploring an unfamiliar dataset
Identifying language that may be missing from an existing concept list
Showing more concepts can reveal nuance, but it can also make the map harder to interpret. Begin broadly, then narrow the view once you identify an area worth investigating.
Explore a topic’s surrounding language
Select or search for a concept to bring its relationships into focus. Stronger relationships become more prominent, while weaker ones fade into the background.
A useful discovery workflow is:
Select a topic you already understand
Examine the concepts positioned closest to it
Select a neighboring concept to continue exploring
Open Records to check how the language is being used
Add useful concepts to a concept list
This is especially helpful when you know the business subject but do not know all the language customers, employees, or respondents use to describe it.
Compare concepts using the axes
Place active concepts on the horizontal and vertical axes to examine how the rest of the map relates to both.
The controls below the map allow Galaxy to position concepts automatically or organize the view around concepts selected for the horizontal and vertical axes.
Fixing concepts to the axes can help you test a relationship question. For example, an organization evaluating its service could:
Fix “service” to one axis.
Fix a descriptive concept, such as “helpful,” “slow,” “frustrating,” or “polite,” to the other axis.
Observe where department or service-area concepts appear.
Replace the descriptive concept with another.
Compare how the department concepts change position.
Watch which concepts:
Sit closer to one axis
Remain between the selected concepts
Sit far from both
Change position when you replace an axis concept
These movements can help you determine which departments, products, or experiences are most strongly associated with a particular description.
Use the Galaxy Concept-Concept Export when you need the numerical association scores behind the visualization. Axis movements and exported scores can help generate and refine hypotheses, but the findings should still be validated using matching records and relevant business context.
Return an axis to Auto when you no longer want to anchor the map to that concept. Use reset to restore the original view.
Find what distinguishes a subset
Unique to this filter is especially powerful in Galaxy because it removes common background language and surfaces concepts that are more characteristic of a selected subset.
For example, you could compare:
One-star and five-star reviews
Launch-period and post-launch feedback
Different products or brands
Regions, departments, or customer groups
Positive and negative business-outlook responses
To use it:
Apply one meaningful filter
Select Unique to this filter
Review the concepts and clusters that appear
Select relevant concepts and inspect their records
Repeat with the comparison group
A concept identified as unique is not necessarily exclusive to that subset. It is more distinctive of the filtered records than of the remaining text analysis project.
If no concepts appear, that can also be informative. It may indicate that the segment uses language that is broadly consistent with the rest of the text analysis project rather than having its own distinctive conversation.
Keep the complete dataset together until you have established this global baseline. If the subset later becomes a separate area of analysis, you can create a focused text analysis project from the matching records in Records.
Use color to tell a clearer story
Galaxy can color concepts by:
Automatically detected conversation clusters
Active concepts
An existing concept list
Conversation-cluster colors are useful for discovery. Concept-list colors are better when you have an established framework or want to present categories consistently.
For example, you might color concepts by product attribute, journey stage, business priority, or SWOT category. Color changes how the map is organized visually; it does not change the underlying associations.
Read more: Use Configure Visualization, Concepts, Filters and Records.
Continue the investigation
Galaxy is most effective for discovery rather than measurement. After finding an interesting relationship:
Use Volume to measure how widely the concepts appear
Use Drivers to examine their association with an outcome
Use Sentiment to understand how they are discussed
Use Records to confirm the relationship in the original language
A strong Galaxy finding explains not only that concepts are close, but also what their relationship suggests and whether the underlying records support that interpretation.