Skip to content

Latest commit

 

History

History
157 lines (109 loc) · 8.02 KB

semantic-search-semantic-text.md

File metadata and controls

157 lines (109 loc) · 8.02 KB
navigation_title mapped_pages applies_to
Semantic search with `semantic_text`
stack serverless

Semantic search with semantic_text [semantic-search-semantic-text]

This tutorial shows you how to use the semantic text feature to perform semantic search on your data.

Semantic text simplifies the {{infer}} workflow by providing {{infer}} at ingestion time and sensible default values automatically. You don’t need to define model related settings and parameters, or create {{infer}} ingest pipelines.

The recommended way to use semantic search in the {{stack}} is following the semantic_text workflow. When you need more control over indexing and query settings, you can still use the complete {{infer}} workflow (refer to this tutorial to review the process).

This tutorial uses the elasticsearch service for demonstration, but you can use any service and their supported models offered by the {{infer-cap}} API.

Requirements [semantic-text-requirements]

This tutorial uses the elasticsearch service for demonstration, which is created automatically as needed. To use the semantic_text field type with an {{infer}} service other than elasticsearch service, you must create an inference endpoint using the Create {{infer}} API.

Create the index mapping [semantic-text-index-mapping]

The mapping of the destination index - the index that contains the embeddings that the inference endpoint will generate based on your input text - must be created. The destination index must have a field with the semantic_text field type to index the output of the used inference endpoint.

PUT semantic-embeddings
{
  "mappings": {
    "properties": {
      "content": { <1>
        "type": "semantic_text" <2>
      }
    }
  }
}
  1. The name of the field to contain the generated embeddings.
  2. The field to contain the embeddings is a semantic_text field. Since no inference_id is provided, the default endpoint .elser-2-elasticsearch for the elasticsearch service is used. To use a different {{infer}} service, you must create an {{infer}} endpoint first using the Create {{infer}} API and then specify it in the semantic_text field mapping using the inference_id parameter.

::::{note} If you’re using web crawlers or connectors to generate indices, you have to update the index mappings for these indices to include the semantic_text field. Once the mapping is updated, you’ll need to run a full web crawl or a full connector sync. This ensures that all existing documents are reprocessed and updated with the new semantic embeddings, enabling semantic search on the updated data.

::::

Load data [semantic-text-load-data]

In this step, you load the data that you later use to create embeddings from it.

Use the msmarco-passagetest2019-top1000 data set, which is a subset of the MS MARCO Passage Ranking data set. It consists of 200 queries, each accompanied by a list of relevant text passages. All unique passages, along with their IDs, have been extracted from that data set and compiled into a tsv file.

Download the file and upload it to your cluster using the Data Visualizer in the {{ml-app}} UI. After your data is analyzed, click Override settings. Under Edit field names*, assign id to the first column and content to the second. Click Apply, then *Import. Name the index test-data, and click Import. After the upload is complete, you will see an index named test-data with 182,469 documents.

Reindex the data [semantic-text-reindex-data]

Create the embeddings from the text by reindexing the data from the test-data index to the semantic-embeddings index. The data in the content field will be reindexed into the content semantic text field of the destination index. The reindexed data will be processed by the {{infer}} endpoint associated with the content semantic text field.

::::{note} This step uses the reindex API to simulate data ingestion. If you are working with data that has already been indexed, rather than using the test-data set, reindexing is required to ensure that the data is processed by the {{infer}} endpoint and the necessary embeddings are generated.

::::

POST _reindex?wait_for_completion=false
{
  "source": {
    "index": "test-data",
    "size": 10 <1>
  },
  "dest": {
    "index": "semantic-embeddings"
  }
}
  1. The default batch size for reindexing is 1000. Reducing size to a smaller number makes the update of the reindexing process quicker which enables you to follow the progress closely and detect errors early.

The call returns a task ID to monitor the progress:

GET _tasks/<task_id>

Reindexing large datasets can take a long time. You can test this workflow using only a subset of the dataset. Do this by cancelling the reindexing process, and only generating embeddings for the subset that was reindexed. The following API request will cancel the reindexing task:

POST _tasks/<task_id>/_cancel

Semantic search [semantic-text-semantic-search]

After the data has been indexed with the embeddings, you can query the data using semantic search. Choose between Query DSL or {{esql}} syntax to execute the query.

::::{tab-set} :group: query-type

:::{tab-item} Query DSL :sync: dsl

The Query DSL approach uses the semantic query type with the semantic_text field:

GET semantic-embeddings/_search
{
  "query": {
    "semantic": {
      "field": "content", <1>
      "query": "What causes muscle soreness after running?" <2>
    }
  }
}
  1. The semantic_text field on which you want to perform the search.
  2. The query text. :::

:::{tab-item} ES|QL :sync: esql

The ES|QL approach uses the match (:) operator, which automatically detects the semantic_text field and performs the search on it. The query uses METADATA _score to sort by _score in descending order.

POST /_query?format=txt
{
  "query": """
    FROM semantic-embeddings METADATA _score <1>
    | WHERE content: "How to avoid muscle soreness while running?" <2>
    | SORT _score DESC <3>
    | LIMIT 1000 <4>
  """
}
  1. The METADATA _score clause is used to return the score of each document
  2. The match (:) operator is used on the content field for standard keyword matching
  3. Sorts by descending score to display the most relevant results first
  4. Limits the results to 1000 documents

::: ::::

Further examples and reading [semantic-text-further-examples]

  • If you want to use semantic_text in hybrid search, refer to this notebook for a step-by-step guide.
  • For more information on how to optimize your ELSER endpoints, refer to the ELSER recommendations section in the model documentation.
  • To learn more about model autoscaling, refer to the trained model autoscaling page.