Skip to content

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

Description

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.

Usage

personalize_create_batch_inference_job(jobName, solutionVersionArn,
  filterArn, numResults, jobInput, jobOutput, roleArn,
  batchInferenceJobConfig, tags, batchInferenceJobMode,
  themeGenerationConfig)

Arguments

  • jobName

    [required] The name of the batch inference job to create.

  • solutionVersionArn

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

  • filterArn

    The ARN of the filter to apply to the batch inference job. For more information on using filters, see Filtering batch recommendations.

  • numResults

    The number of recommendations to retrieve.

  • jobInput

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

  • jobOutput

    [required] The path to the Amazon S3 bucket where the job's output will be stored.

  • roleArn

    [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

    The configuration details of a batch inference job.

  • tags

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

  • batchInferenceJobMode

    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

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

Value

A list with the following syntax:

list(
  batchInferenceJobArn = "string"
)

Request syntax

svc$create_batch_inference_job(
  jobName = "string",
  solutionVersionArn = "string",
  filterArn = "string",
  numResults = 123,
  jobInput = list(
    s3DataSource = list(
      path = "string",
      kmsKeyArn = "string"
    )
  ),
  jobOutput = list(
    s3DataDestination = list(
      path = "string",
      kmsKeyArn = "string"
    )
  ),
  roleArn = "string",
  batchInferenceJobConfig = list(
    itemExplorationConfig = list(
      "string"
    ),
    rankingInfluence = list(
      123.0
    )
  ),
  tags = list(
    list(
      tagKey = "string",
      tagValue = "string"
    )
  ),
  batchInferenceJobMode = "BATCH_INFERENCE"|"THEME_GENERATION",
  themeGenerationConfig = list(
    fieldsForThemeGeneration = list(
      itemName = "string"
    )
  )
)