Attributes
When you import your data into Zoovu's Data Platform, it is turned into a powerful ontology consisting of attributes - fundamental elements used to organize and manage product data.
View and manage product attributes
Once your product data is uploaded and processed, you can find the product attributes listed under Data Platform > Ontology > Attributes.
In the top right corner, you'll see a switch that lets you toggle between two views:
Your attributes: This view shows attributes that were either automatically created from your data feed or manually created by you. Zoovu attributes: The attributes extracted from your data based on Zoovu’s ontology.
Attribute types
| Attribute type | Description | Example |
|---|---|---|
| Textual | Textual concepts represent attributes with predefined values | color |
| Numeric | Numeric concepts represent attributes with a number value and can also define units | weight |
| Boolean | Boolean concepts represent attributes with either a true or false value | true |
| Untyped | Untyped concepts represent attributes with arbitrary values. Values of untyped attributes are not reflected in concepts themselves | |
| Category | Category concepts are used to classify products. Each category represent a node in the category tree. | Power tools |
You can change attribute types. For example, if your attribute "Rating" is text-based (e.g., "5 stars"), but you need it in numeric form (e.g., "5.0"), you can change it from a text attribute to a numeric attribute.
Extraction rules
Extraction rules determine when an attribute should be extracted and which values should be captured. The platform automatically generates extraction rules based on your incoming data but you can also edit them manually.
Select any attribute in Ontology > Attributes and go to the "Extraction Rules" tab to modify the rules:
Examples of extraction rules:
- Matches: The data must exactly match the rule. For example, the "Color" attribute must exactly match "Black."
- Phrase match: The rule matches an entire phrase, like "Battery Life" matching "up to 10 hours."
- Contains: The rule matches when the input contains the specified phrase, e.g. extracting "16GB" from "16GB RAM."
- Regular expression: Allows for complex matching, e.g. different formats of model numbers.