Learn about the different formats available when exporting your analyzed data.

It may be necessary to export your results in order to create custom reports or visualizations using third party tools.

There are two types of exports available: a detailed export (including metadata and all analytics) and an export of the current Filterset. We will go over both below.

It is important to keep in mind that the filters that are applied when the export is created will be reflected in both types of exports.

Exporting your results

Click "Export to CSV" to reveal the export options.

You have a choice of a detailed export – "Export metadata and all analytics" – or a lighter version – "Export current Filterset". Remember that for both options all filters will be applied to the export.

All exports are downloaded as a compressed file that contains the requested export in CSV format. If the results being exported come from different sources, each source will be a separate file.

Comparing the two exports

To demonstrate, here we have exported a filtered subset of our data that contains results related to the Topic "room". We’ll look at a sample of how this analysis has been translated into the different export types.

Since we applied a Topic filter, only the results that include the Topic "room" will be exported. If you'd like to export all of the unfiltered results, ensure there are no active filters of any kind. This applies for both types of exports.

Export metadata and all analytics

The first few column headers in this export type mirror those in the CSV you uploaded, and these are followed by columns containing the detected analytics. This type of export is ideal for reconstructing the full analysis because it’s a flat model of the analytics results displayed in the application. The unique IDs map out the links between the data.

Here's an overview of the exported analytics. We’ve used Excel to show the CSV screenshots.

statement_expression - Reproduces the section from the original record in which the result appears.

language - The detected language for this record is indicated through the language two-letter ISO code.

statement_topic - Shows the identified Topic in its exact form, as mentioned by the customer.

statement_opinion - Shows the identified Opinion in its exact form, as mentioned by the customer.

group_unique_doc_count - Displays the number of records in which this statement and all its variations show up.

sentiment - Identifies the sentiment expressed by the customer (Praise, Problem, Suggestion, Question, or Mention).

group_topic - Is the Topic representative for that group, meaning the group is displayed using this term. Read more about groups here.

group_opinion - Is the Opinion representative for that group, meaning the group is displayed using this term. Read more about groups here.

tags - Any tags applied to the Comment are listed in this column.

Export Current Filterset

This export type gives a very high-level overview of the Analysis page you are currently looking at. It differs depending on the way the results are displayed - as Comments, Topics, or Opinions. This format is ideal for simple graphs and charts.

Export Current Filterset - as Comments

expression - Reproduces the section from the original record in which the Comment appears.

topic - Is the Topic representative for that group, meaning the group is displayed using this term. Read more about groups here.

opinion - Is the Opinion representative for that group, meaning the group is displayed using this term. Read more about groups here.

sentiment - Identifies the sentiment expressed by the customer (Praise, Problem, Suggestion, Question, or Mention).

record_count - A count of all the Records where this Comment appears.

tags - Any tags applied to the Comment are listed in this column.

Export Current Filterset - as Topics

topic - Shows the Topic. Since we originally applied a filter to see results containing the Topic “room”, we only have one Topic.

praises - A count of all the instances where “room” appears as a Topic in a Praise.

problems - A count of all the instances where "room" appears as a Topic in a Problem.

suggestions - A count of all the instances where “room" appears as a Topic in a Suggestion.

questions - A count of all the instances where “room" appears as a Topic in a Question.

mentions - A count of all the instances where “room" appears as a Topic in a Mention.

total - A sum of the previous 5 columns.

tags - Any tags applied to the Topic are listed in this column.

Export Current Filterset - as Opinions

opinion - The Opinion associated with the Topic "room" is indicated in this column.

praises - A count of all the instances where the Opinion appears as a Praise.

problems - A count of all the instances where the Opinion appears as a Problem.

questions - A count of all the instances where the Opinion appears as a Question.

suggestions - A count of all the instances where the Opinion appears as a Suggestion.

mentions - A one word Opinion will never be classified as a Mention - it will fall under Praise, Problem, Suggestion, or Question. For example: “Great” (Praise), “Gross” (Problem), “Why?” (Question), or “Should be” (Suggestion). Therefore when exporting Opinions the mentions column will always be 0.

total - A sum of the previous 5 columns.

tags - Any tags applied to the Opinion are listed in this column.

Using Exports for Visualizations

Exporting your results as a CSV can be very useful for making visualizations. While advanced users might use Tableau to make interactive visualizations, any user can get great results out of Google Studio, Excel or a simple word cloud tool.

Here are some example graphs made with Keatext exported data and Tableau by a professional designer.

Did this answer your question?