How to do it...

We want use the feature type to implement a common PageRank scenario where documents are scored base of same characteristics. To achieve this will execute the following steps:

1. To be able to score base on a pagerank value and an inverse url length, we will use a similar mapping:

PUT test-feature
{
  "mappings": {
    "properties": {
      "pagerank": {
        "type": "feature"
      },
      "url_length": {
        "type": "feature",
        "positive_score_impact": false
      }
    }
  }
}

1. Now, we can store a document as shown here:

PUT test-feature/_doc/1
{
  "pagerank": 5,
  "url_length": 20
}

1. Now, we can execute a feature query on the pagerank value to return our record with a similar query, like so:

GET test-feature/_search
{
  "query": {
    "feature": {
      "field": "pagerank"
    }
  }
}

1. The evolution of the previous feature functionality is to define a vector of values using the feature_vector type; usually it cab be use to score by topics, categories, or similar discerning facets. We can implement this functionality using the following steps:

So, the following code defines the mapping for the categories field:

PUT test-features
{
  "mappings": {
    "properties": {
      "categories": {
        "type": "feature_vector"
      }
    }
  }
}

1. We can now store some documents in the index by using the following commands:

PUT test-features/_doc/1
{
  "categories": {
    "sport": 14.2,
    "economic": 24.3
  }
}

PUT test-features/_doc/2
{
  "categories": {
    "sport": 19.2,
    "economic": 23.1
  }
}

1. Now, we can search based on saved feature values, as shown here:

GET test-features/_search
{
  "query": {
    "feature": {
      "field": "categories.sport"
    }
  }
}