Record Filters (filters that are specific to a dataset) are based on the data you upload. They mirror the columns in your CSV file. Record Filters work best with the structured data that accompanies the feedback. Record filters are a powerful component of text mining and can be used for segmentation.

If your dataset only contains a comment/feedback column, use the Search Filter instead. It offers the same functionality as the Record Filter, but with additional control and usability.

On the left, you will see the Filter panel. If you can't see it, click on "Filters" in the upper left-hand corner to open it.

To apply a Record Filter, select + Add field filter. You will then need to choose the Source (dataset) you wish to filter from in the dropdown menu that appears. Only the Source that was selected in the Analyzed Sources section of the panel will appear in this dropdown. In this example we only have one Source available - Reviews.

Each column in the original dataset acts as a Record Filter. Select the column you wish to filter by. The type of filter (text, numbers or dates) is dictated by the nature of the dataset before the upload.

You can apply multiple Record Filters at the same time. If you apply multiple filter parameters within the same Record Filter (column), OR Boolean logic is used, with the exception of number data types which uses AND logic. If multiple filter parameters are chosen amongst different Record Filters (columns), AND Boolean logic is used.