Why do we need groups? 

When people provide feedback, it’s highly likely that their comments use a large variety of words and phrases to express the same idea. Our users want to see a clear prioritization based on how widespread a specific Problem or Praise is. The AI technology behind the Keatext platform groups all these variations automatically to allow the user to save time. 

The above example shows variations in the Indicators. “Dirty”, “filthy”, “stained”, “grubby”, “gross”, “not been cleaned” and even the misspelling “unclen” are all grouped. Each group has a representative label, which in this case is “dirty”. The labels are chosen automatically based on frequency and other parameters, such as length. The user has control over both Topic groups and Indicator groups, which in turn affect how Issues are grouped. 

A group is a container for one or several Topics and Indicators that are synonymous, nearly synonymous or should be treated as such by Keatext.

Why customize groups?

In some instances users may want to customize the analysis to further adapt and refine the results to suit their needs. There are several reasons why a user would choose to do this, including to reflect domain knowledge in the results, change the focus of the analysis or tweak some visible errors. Groups can also reflect organizational knowledge—for example, a unit and an apartment may mean the same thing for an apartment leasing company. Your organization may also use abbreviations (such as CSR for customer service representative in call center transcripts or other documents.)

Can I use grouping to improve accuracy? 

For small tweaks, modifying groups can be quite powerful. However, for recurring issues with accuracy, it’s best to discuss with the account manager to see if a retraining is a better option than groups. 

Are the changes permanent?

The user changes to groups can be reset. However, it’s important to note that if several users are part of the same organization, the changes will happen for all users at an organizational level. The reset button returns the groups to default, so it’s best avoided if your organization has spent hours customizing its groups. 

How do I merge specific Topics or Indicators into a group?

If you wish to combine two or more Topics or Indicators because they represent the same concept, click on the Topic or Indicator display icon at the top right hand of your screen.

Once in the Topic or Indicator display, click on the selection box next to one or more results. This reveals the Merge functionality in a menu above the results list. 

In the screenshot above, I have selected "Place" and "Area" since in the context of this dataset they express the same concept. After clicking Merge, you can choose to add the selected results to an existing group or create a new one. 

How do I manage groups? 

First, head to the group management page by  clicking on the management icon in the navigation menu on the left of the screen. This page is shared by watchlists and groups. To manage groups, click on either Group Topics or Group Indicators.

Use search to find the group you’re looking for. 

Groups can be managed either by modifying the settings for that group or in bulk by selecting the boxes in front of the groups. Here are the options:

Disable Group - This is a temporary way of deleting a group. Try this out before deciding to dissolve a group. Each member will be part of an individual group. 

Enable Group - Undo Disable Group by returning the members to this group. 

Merge Group - Combine groups into bigger groups.   

Dissolve Group - Permanently deletes a group. Make sure everyone in the organization is ok with this. 

In the upper right corner you’ll also see the “Reset Groups” button. This resets all the groups, including groups that were formed automatically for your dataset. This can dramatically change the results of your analysis. 

Click on a single group if you want to inspect each member. Members can also be managed by removing them from the group or transferring them to other groups. 

Did this answer your question?