Use Vector Filters in Data 360 Segments (Beta)

Vector filters reduce reliance on exact keyword matches and manual data organization, allowing you to create accurate segments efficiently. This feature uses NLP algorithms to sort through unstructured data in segment creation. It supports filtering both structured and unstructured data within a single interface and a segment.

Required Editions

Note Vector filters in Segments is a beta service that is subject to the Beta Services Terms at Agreements - Salesforce.com or a written Unified Pilot Agreement if executed by Customer, and applicable terms in the Product Terms Directory. Use of this beta service is at the Customer's sole discretion.

Available in: All Editions supported by Data 360. See Data 360 edition availability.

Permission Sets Needed
To use vector filters:	Any one of these permission sets: Data Cloud Architect Data Cloud Activation Manager Data Cloud Activation Specialist

1. In Data Cloud, click Segments.
2. Select the segment from the list.
3. Click Edit Rules.
4. Click the Include or Exclude tab. You can have up to 50 filters in each tab.
5. You can add criteria to:
   • Direct Attributes‌ — a single data point or piece of information about the entity being segmented, such as a postal code, first name, or birthday. Direct attributes are often demographic attributes.
   • Related Attributes‌ ‌— a collection of multiple data points or pieces of information associated with the entity being segmented, such as purchase history, product numbers, or email interactions. Related attributes are often behavioral or engagement events.
   • Calculated Insights ‌— ‌search for calculated insights by name or find them under the object referenced in the SQL. You can only use processed calculated insights in a segment.
6. Select the Is Similar To operator.
7. Enter the required value. For example, backpacks.
8. Select the similarity level using the slider. By default, it’s set to 100.
9. Click Compute Results.
10. Click View Results. You'll see these details:
    • The number of similar terms returned based on the search term and the selected similarity level.
    • A sample of the five most similar terms.
    • A sample of the five least similar terms.

    You can compute results with different similarity levels and different values as needed.

11. Click Done to save the filter.

The filter operates within its defined scope, making sure the data it collects is relevant and fits the context you're looking for, thus improving the accuracy of your segment data.