Sort Discover results by relevance

This example uses the web logs sample data, or you can use your own data.

1. In Discover, make sure you're in ES|QL mode. If Discover opens in classic mode, refer to Switch between ES|QL and classic mode.

2. To get scores, add METADATA _score to the FROM command, and match text with a full-text function such as MATCH. Then sort by _score. For the sample data, enter:

   FROM kibana_sample_data_logs METADATA _score
   | WHERE MATCH(message, "Mozilla")
   | SORT _score DESC
   | KEEP _score, message
   | LIMIT 10
   

3. If you don't see any results, expand the time range, for example to Last 7 days.

Result: The table lists the best matches first, with their scores in the _score column.

For more about scores, refer to Relevance scoring and to the _score metadata field.

This example uses the flights sample data, or you can use your own data.

1. In Discover, if it opens in ES|QL mode, switch to classic mode. Then, from the data view menu, select the data that you want to work with.

   For the flights sample data, set the data view to Kibana Sample Data Flights.

2. Run your search. For the sample data, try:

   Warsaw OR Venice OR Clear
   

3. If you don't see any results, expand the time range, for example to Last 7 days.

4. In the fields list in the sidebar, expand the Meta fields list. Next to _score, select (Add "_score" field) to add it as a column.

5. Optional: Add any other fields you want to the document table.

6. By default, the table sorts by the time field of your data view, which is timestamp in the flights sample data. To turn off that sorting, select the field sorted option, and then select Clear sorting.

7. Open the Pick fields to sort by menu, and then select _score.

8. Select High-Low.

   
   

Result: The table sorts documents from most to least relevant.