Skip to main content
Version: 1.16

REGARDS Meta catalog

Meta catalog

REGARDS Meta catalog is the result of the standardization of multiples source catalogs of products in a single catalog.

In all REGARDS documentation, we will talk about :

  • Data source for an existing catalog crawled by REGARDS.
  • Meta catalog for the global resulting catalog containing all products from all data sources crawled.

To learn more, you can read Functional Overview - Meta catalog services.

Model

The notion of Model is a key notion and defines the list of metadata for a group of valid products. It is used to standardize products.

info

REGARDS Meta catalog can contain products with different Models.

Data source

Products of each crawled data source are transformed by REGARDS to match a Model before being inserted in the final catalog called Meta catalog.

Every Data source is associated to a Model to ensure every product is valid and contains associated metadata.

Product standardization

Thanks to Model and Data source design, you can transform products coming from several data sources to standardize them into the same model. This is the key concept in REGARDS that allows you to search products coming fom several providers with the same search criteria.

Let's take an example

Data source 1 products example
{
"label": "my product from source 1",
...
"properties": {
"creation_date": "2024-01-01"
}
}
Data source 2 products example
{
"description": "my product from source 2",
"date": "2023-06-01"
}

In this example, you can define a Model to standardize products coming from these two different datasource, which use 2 different products formats, to make products sharing the same standard Model. To do so, you can take these two metadata attributes, date and properties.creation_date, to set them in the same attribut in REGARDS Meta catalog.
With those two attributes indexed with the same attribute in the resulting meta catalog, you can search for both products with the same search criteria.