Importing Data

You can create a Data Import Job to load interaction data to the system.

To do so, prepare an input data file (CSV). Use meaningful names for data columns so that you can map them to interaction attributes. You can map multiple columns of the input file to each attribute.

Use case for Data Import

Let us assume that you are having an interaction with multiple attributes. For example, Date, Website, Email, Location, and so on. These attributes may have more than one value. Such that there can be two dates, three websites, two locations, and so on.

You can add these data to the CSV file (with meaningful columns such as Date1, Date2, Web1, Web2). Once the file is imported, you can then map Date1 and Date2 to the corresponding Date attributes, Web1 and Web2 to the corresponding Website attributes, and so on.

Steps to Import Data

The imported data is displayed on the Data Import page. The main content of the imported file is split into two parts:

  • COLUMNS - on the left side, as per the columns found in the imported file.
  • ATTRIBUTES/MEMBERS/CROSSWALKS - tabs on the right side. The imported file contains columns for attributes, members, and crosswalks. You can toggle between the three tabs and map the data.

Using the multi-mapping feature, you can map columns from the imported file (Columns appearing on the left side of the Data Import page) to attributes, members, or crosswalks.

Perform the following steps to map the imported file to attributes, members, and crosswalks:

  1. Navigate to the Jobs page of Data Science and click CREATE NEW JOB.
  2. On the Create new job page, enter the Job name and Cluster size.

  3. Click ADD NEW TASK. In the pop-up window that appears, click Data Import.

  4. In the Data settings pop-up window that appears, enter the following information:
    1. AWS Key and AWS Secret - for accounts where the import data file is stored
    2. S3 bucket name - indicates the bucket where the file is stored
    3. Data file - indicates the path to the file
    4. Mapping file - indicates the file that maps columns from the data file to the interaction type attributes
    5. Interaction type - interaction types are taken from the tenant’s business configuration. In this pop-up window, they appear in a list. Click the drop-down arrow and choose the one for which you need to load data.

  5. Once you have entered all the values, click DONE. The Data Import page appears, where you can view the values that you entered.

  6. You now have to map the Columns listed on the left to the Attributes (under the ATTRIBUTES tab on the right side). Perform the following steps to complete the mapping:
    1. On the left side, under Columns, select a particular column.

    2. To map an attribute with the column that you have selected, click its corresponding check box.

    3. You can map multiple columns to any attribute, except for nested attributes. This is how the mapping behaves:
      • Arrows appear for the currently selected element.
      • The linked column appears next to its linked attribute. And the linked attribute appears next to its linked column. To remove the link, click X.

  7. Move to the MEMBERS tab to map the columns to member source type and IDs. Perform the following steps to complete the mapping:
    1. To map the source type, select the column from the combo-box.
    2. To map the ID, select the column from the Columns list and then click the check box for the member.

  8. Click ADD CROSSWALK to add more crosswalks to any member.

  9. Move to the CROSSWALKS tabs. Map the source type and ID the same way as you did under the MEMBERS tabs.
  10. Once you have mapped the attributes, members, and crosswalks, click DONE.
  11. The Data Import task appears as a list, under the Tasks section. Click EDIT TASK to further modify the task. Following are some additional options available:
    1. Inactive - when enabled, the interactions are stored only into S3 and in Cassandra
    2. Dry run - when enabled, the task can be tested without saving
    3. Stop on failure - when enabled, the job process stops on detecting an error

  12. When ready, click RUN. The Jobs page opens with the recently imported job added to the existing list.