Is there a better way to perform a "range sort" or "bucket sort" query?

Elastic Stack Elasticsearch
1

Hi, we're using ES 5.5.2, say we got this scenario:

One user searched "Christmas presents", and the results were:

Christmas present A _score: 35.12, price: 100
Christmas present B _score: 34.56, price: 120
Christmas present C _score: 34.33, price: 100
Valentine present A _score: 12.11, price: 80
Valentine present B _score: 12.02, price: 80
Valentine present C _score: 11.45, price: 70

If the user clicked "sort by price asc", the first page's results may be not relevant enough.
We suppose it is better to sort price in each _score bucket [0, 10), [10, 20), [20, 30), ...

So we changed the search query to:

{
  "from": 0,
  "aggs": {},
  "query": {
    # .....
  },
  "sort": {
    "_script": {
      "type": "number",
      "script": {
        "lang": "painless",
        "inline": "doc['price'].value + Math.floor(_score / 10.0) * 100000"
      },
      "order": "asc"
    }
  },
  "track_scores": true
}

Or is there a better way to perform like this? Thank you for help!

2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.