Personalize / Client / create_batch_inference_job

create_batch_inference_job#

Personalize.Client.create_batch_inference_job(**kwargs)#

Generates batch recommendations based on a list of items or users stored in Amazon S3 and exports the recommendations to an Amazon S3 bucket.

To generate batch recommendations, specify the ARN of a solution version and an Amazon S3 URI for the input and output data. For user personalization, popular items, and personalized ranking solutions, the batch inference job generates a list of recommended items for each user ID in the input file. For related items solutions, the job generates a list of recommended items for each item ID in the input file.

For more information, see Creating a batch inference job.

If you use the Similar-Items recipe, Amazon Personalize can add descriptive themes to batch recommendations. To generate themes, set the job’s mode to THEME_GENERATION and specify the name of the field that contains item names in the input data.

For more information about generating themes, see Batch recommendations with themes from Content Generator.

You can’t get batch recommendations with the Trending-Now or Next-Best-Action recipes.

See also: AWS API Documentation

Request Syntax

response = client.create_batch_inference_job(
    jobName='string',
    solutionVersionArn='string',
    filterArn='string',
    numResults=123,
    jobInput={
        's3DataSource': {
            'path': 'string',
            'kmsKeyArn': 'string'
        }
    },
    jobOutput={
        's3DataDestination': {
            'path': 'string',
            'kmsKeyArn': 'string'
        }
    },
    roleArn='string',
    batchInferenceJobConfig={
        'itemExplorationConfig': {
            'string': 'string'
        }
    },
    tags=[
        {
            'tagKey': 'string',
            'tagValue': 'string'
        },
    ],
    batchInferenceJobMode='BATCH_INFERENCE'|'THEME_GENERATION',
    themeGenerationConfig={
        'fieldsForThemeGeneration': {
            'itemName': 'string'
        }
    }
)
Parameters:

  • jobName (string) –

    [REQUIRED]

    The name of the batch inference job to create.

  • solutionVersionArn (string) –

    [REQUIRED]

    The Amazon Resource Name (ARN) of the solution version that will be used to generate the batch inference recommendations.

  • filterArn (string) – The ARN of the filter to apply to the batch inference job. For more information on using filters, see Filtering batch recommendations.

  • numResults (integer) – The number of recommendations to retrieve.

  • jobInput (dict) –

    [REQUIRED]

    The Amazon S3 path that leads to the input file to base your recommendations on. The input material must be in JSON format.

    • s3DataSource (dict) – [REQUIRED]

      The URI of the Amazon S3 location that contains your input data. The Amazon S3 bucket must be in the same region as the API endpoint you are calling.

      • path (string) – [REQUIRED]

        The file path of the Amazon S3 bucket.

      • kmsKeyArn (string) –

        The Amazon Resource Name (ARN) of the Key Management Service (KMS) key that Amazon Personalize uses to encrypt or decrypt the input and output files.

  • jobOutput (dict) –

    [REQUIRED]

    The path to the Amazon S3 bucket where the job’s output will be stored.

    • s3DataDestination (dict) – [REQUIRED]

      Information on the Amazon S3 bucket in which the batch inference job’s output is stored.

      • path (string) – [REQUIRED]

        The file path of the Amazon S3 bucket.

      • kmsKeyArn (string) –

        The Amazon Resource Name (ARN) of the Key Management Service (KMS) key that Amazon Personalize uses to encrypt or decrypt the input and output files.

  • roleArn (string) –

    [REQUIRED]

    The ARN of the Amazon Identity and Access Management role that has permissions to read and write to your input and output Amazon S3 buckets respectively.

  • batchInferenceJobConfig (dict) –

    The configuration details of a batch inference job.

    • itemExplorationConfig (dict) –

      A string to string map specifying the exploration configuration hyperparameters, including explorationWeight and explorationItemAgeCutOff, you want to use to configure the amount of item exploration Amazon Personalize uses when recommending items. See User-Personalization.

      • (string) –

        • (string) –

  • tags (list) –

    A list of tags to apply to the batch inference job.

    • (dict) –

      The optional metadata that you apply to resources to help you categorize and organize them. Each tag consists of a key and an optional value, both of which you define. For more information see Tagging Amazon Personalize resources.

      • tagKey (string) – [REQUIRED]

        One part of a key-value pair that makes up a tag. A key is a general label that acts like a category for more specific tag values.

      • tagValue (string) – [REQUIRED]

        The optional part of a key-value pair that makes up a tag. A value acts as a descriptor within a tag category (key).

  • batchInferenceJobMode (string) –

    The mode of the batch inference job. To generate descriptive themes for groups of similar items, set the job mode to THEME_GENERATION. If you don’t want to generate themes, use the default BATCH_INFERENCE.

    When you get batch recommendations with themes, you will incur additional costs. For more information, see Amazon Personalize pricing.

  • themeGenerationConfig (dict) –

    For theme generation jobs, specify the name of the column in your Items dataset that contains each item’s name.

    • fieldsForThemeGeneration (dict) – [REQUIRED]

      Fields used to generate descriptive themes for a batch inference job.

      • itemName (string) – [REQUIRED]

        The name of the Items dataset column that stores the name of each item in the dataset.

Return type:

dict

Returns:

Response Syntax

{
    'batchInferenceJobArn': 'string'
}

Response Structure

  • (dict) –

    • batchInferenceJobArn (string) –

      The ARN of the batch inference job.

Exceptions