Amazon Personalize workflow details

Match your use case to Amazon Personalize resources – Amazon Personalize features domain based resources and custom resources configured for different cases. When you match your use case to an Amazon Personalize resource, note its data requirements. After you choose a use case or recipe, this information can help as you prepare your data.

Prepare your training data – Based on your domain use case or custom recipe's data requirements, prepare your bulk training data in a CSV file. Depending on your use case or recipe, Amazon Personalize can use item interaction, item, user, action, and action interaction data. If you don't have bulk data, you can use individual import operations to collect data and stream events until you meet Amazon Personalize training requirements and the data requirements of your domain use case or recipe.

Create schema JSON files for your data – Create schema JSON files for each type of data that you are importing. These files outline the structure and content of your data, including column names and their data types.

Create a dataset group – A dataset group is a container for Amazon Personalize resources. You can create a Domain dataset group with preconfigured resources for VIDEO_ON_DEMAND or ECOMMERCE domains. Or you can create a Custom dataset group and create only custom resources.

Create schemas and datasets – A schema tells Amazon Personalize about the structure of your data and allows Amazon Personalize to parse the data. A dataset is a container for training data in Amazon Personalize.

Import training data into datasets – Import your prepared interaction, item, user, action, or action interaction records. You can import records in bulk or individually.

Train and deploy a model – To train and deploy a model for at the VIDEO_ON_DEMAND or ECOMMERCE domains, you create domain recommenders. For custom resources, you create a custom solution and a solution version. For real-time recommendations, you deploy the solution version in a campaign.

Get recommendations – Use your recommender or custom campaign to get recommendations. You can use filters to include or exclude certain types of items from recommendations. For more information, see Filtering recommendations and user segments. With custom resources, you can also get batch recommendations or user segments without creating a campaign.

Record real-time events – Record real-time events as your customers interact with recommendations. This builds out your interactions data and keeps your data fresh. And it tells Amazon Personalize about the current interests of your user, which can improve recommendation relevance.