Rank feature datatype

A rank_feature field can index numbers so that they can later be used to boost documents in queries with a rank_feature query.

PUT my_index
{
  "mappings": {
    "properties": {
      "pagerank": {
        "type": "rank_feature" 
      },
      "url_length": {
        "type": "rank_feature",
        "positive_score_impact": false 
      }
    }
  }
}

PUT my_index/_doc/1
{
  "pagerank": 8,
  "url_length": 22
}

GET my_index/_search
{
  "query": {
    "rank_feature": {
      "field": "pagerank"
    }
  }
}

Rank feature fields must use the rank_feature field type

Rank features that correlate negatively with the score need to declare it

rank_feature fields only support single-valued fields and strictly positive values. Multi-valued fields and negative values will be rejected.

rank_feature fields do not support querying, sorting or aggregating. They may only be used within rank_feature queries.

rank_feature fields only preserve 9 significant bits for the precision, which translates to a relative error of about 0.4%.

Rank features that correlate negatively with the score should set positive_score_impact to false (defaults to true). This will be used by the rank_feature query to modify the scoring formula in such a way that the score decreases with the value of the feature instead of increasing. For instance in web search, the url length is a commonly used feature which correlates negatively with scores.