1. Create a Dataset

Before you can start a job, you need to create a dataset. You can do that in the App by clicking the New Dataset button. Right now the application will only accept CSV files.

Add a dataset

Add a dataset

Your dataset should only include the fields relevant for the task (this can be multiple fields), you can either edit the dataset before uploading or enter the specific Columns in the upload form.

Select specific fields from the dataset

Select Columns

2. Start a Ca Job

Once your datasets have been created, it's time to kick off a job. You can do that in the App by clicking the Start Job button.

Start a categorize job

Start Categorize Job

  1. Select New Job
  2. Select Categorize
  3. Enter a name for your job - this can be anything you want
  4. Select the Dataset you added in the previous step from the dropdown
  5. In the Categories section, click the Auto Determine button. This will open up a panel where you can provide instructions and the system can automatically generate clusters for you. Alternatively, you can enter the specific clusters for your job in the Categories input field. You can enter multiple clusters at once by separating them with a | character.
  6. If you want to give the app specific guidance when determining which record belongs in which cluster(s), click Edit on the Instructions box. While not strictly necessary, this can help guide the app to cluster your data more accurately.

3. Exporting Your Clusters

Since a job can take a while depending on the amount of data, you can see the progress on the job’s page which will update automatically ever 30 seconds. You will also receive an email when the job is complete, so feel free to navigate away from the page if needed.

View Job Progress

View Job Progress

Once the categorize job is complete, you can click the Export Data button to get the corresponding cluster values in a CSV file.