Datasets page

User permissions required: ‘View Sources’.

Datasets overview page

After signing in, you’ll be presented with the Datasets overview page.

Alternatively, you can navigate to this page anytime by clicking the 'UiPath Communications Mining' logo in the top left of your page.

From this page you can:

• See all the Datasets you have access to
• Edit or delete these Datasets (User permissions required: ‘Datasets admin’)
• Navigate to other pages in the platform

You can navigate straight into a dataset by clicking one of the three options (Explore, Train, and Reports) listed underneath it.

When looking at the datasets you have access to, you can filter using the drop-down menu to a specific project that you are a member of to restrict the number of datasets on display.

You can also search for a specific dataset by name using the search bar.

Datasets page

Selecting a dataset

Each dataset card gives you some useful information on the dataset:

Example dataset card

Each dataset card references:

• The dataset title and description
• The project the dataset is link to and the dataset name (project/name)
• The sources connected to the dataset
• The model family (language)
• If sentiment analysis is enabled
• When the dataset was last changed (and when it was created on hover)

Select Explore, Train and Reports beneath the dataset information card, to navigate to those pages.

Copy an existing dataset

• Select the Duplicate option if you want to copy an existing dataset from another dataset (this will auto-select the same sources, and sentiment selection as that dataset)

• Select all the (additional) sources that you want to connect to the dataset

What does copying a dataset mean and why would you do it?

When you create a new dataset, you can choose to essentially create a carbon copy of a pre-existing dataset. This means that you copy over the same sources, entities, sentiment selection, labels and reviewed examples as the dataset you've copied the taxonomy from.

You can then work on the copy dataset (which will require a different name) and make changes to it freely without impacting the original.

There are two main reasons why you want to do this:

1. You want to make major changes to your model, in terms of dataset structure for instance, and want to preserve the original dataset in case you want to revert back to it
2. You want to use the work already done by annotating the original dataset and create a new dataset to which you can add additional sources of a similar nature.

Dataset settings page

As well as the Datasets overview page, each dataset has its own individual settings page. This can be accessed by clicking into the dataset and going to 'Settings'.

A dataset’s settings page contains useful information about the dataset and is where you can perform various actions.

The page is split into three tabs:

• General - where you can see annotating statistics and the message metadata properties
• Labels & Entities - where you can add, update, create and disable entities, as well as add label descriptions and see and upload or download their taxonomy in full
• Settings - where you can update the global settings of the dataset, including title, description and sources

From the datasets page, if you have the Datasets admin permission, you can also delete the dataset using the delete button on the bottom of the 'General' tab. This will only appear if you have the required permission.

Default Dataset settings page