Adaptive search result re-ranking
A system and method to provide adaptively re-ranked search results in response to a search query. An initial ranking of search results associated with a search query is established. A query identifier associated with the search query received from an end-user system is identified. Based on the query identifier, a set of model parameters associated with the search query are identified. In response to receiving the search query, using a machine-learning model, a re-ranking of the search results based at least in part on the set of model parameters and the initial ranking of the search results is determined. The re-ranked search results are provided to the end-user system.
1. A method comprising:
establishing an initial ranking of search results associated with a search query cluster comprising a first search query and a second search query;
identifying, by a processing device, a first query identifier associated with the first search query received from a first end-user system;
identifying, based on the first query identifier, a set of model parameters associated with the first search query;
in response to receiving the first search query, generating, using a machine-learning model, one or more changes to configure a re-ranking of the search results based at least in part on the set of model parameters and the initial ranking of the search results;
associating the re-ranking of the search results with the search query cluster;
generating a first graphical user interface to display the re-ranking of the search results to the first end-user system;
executing, by the machine-learning model, a test using the one or more changes to generate a set of adjusted model parameters associated with the search query cluster;
identifying a second query identifier associated with the second search query received from a second end-user system; and
generating a second graphical user interface to display the re-ranking of the search results to the second end-user system.
2. The method of claim 1 , wherein the set of model parameters comprises an attractiveness score associated with each of the set of search results corresponding to the search query cluster.
3. The method of claim 1 , further comprising swapping a position of a first search result and a second search result of the re-ranking of the search results to generate a further re-ranking of the search results.
4. The method of claim 3 , further comprising in response to receiving a subsequent submission of the first search query, causing the further re-ranking of the search results to be provided.
5. The method of claim 1 , wherein the set of adjusted model parameters is generated based at least in part on data points associated with search events collected over a period of time.
6. The method of claim 1 , further comprising generating subsequent re-rankings of the search results based on the set of adjusted model parameters.
7. A system comprising:
a memory to store instructions; and
a processing device operatively coupled to the memory, the processing device to execute the instructions to perform operations comprising:
establishing an initial ranking of search results associated with a search query cluster comprising a first search query and a second search query;
identifying a first query identifier associated with the first search query received from a first end-user system;
identifying, based on the first query identifier, a set of model parameters associated with the first search query;
in response to receiving the first search query, generating, using a machine-learning model, one or more changes to configure a re-ranking of the search results based at least in part on the set of model parameters and the initial ranking of the search results;
associating the re-ranking of the search results with the search query cluster,
generating a first graphical user interface to display the re-ranking of the search results to the first end-user system;
executing, by the machine-learning model, a test using the one or more changes to generate a set of adjusted model parameters associated with the search query cluster;
identifying a second query identifier associated with the second search query received from a second end-user system; and
generating a second graphical user interface to display the re-ranking of the search results to the second end-user system.
8. The system of claim 7 , wherein the set of model parameters comprises an attractiveness score associated with each of the set of search results corresponding to the search query cluster.
9. The system of claim 8 , the operations further comprising swapping a position of a first search result and a second search result of the re-ranking of the search results to generate a further re-ranking of the search results.
10. The system of claim 9 , the operations further comprising, in response to receiving a subsequent submission of the first search query, causing the further re-ranking of the search results to be provided.
11. The system of claim 7 , wherein the set of adjusted model parameters is generated based at least in part on data points associated with search events collected over a period of time.
12. The system of claim 7 , the operations further comprising generating subsequent re-rankings of the search results based on the set of adjusted model parameters.
13. A non-transitory computer readable storage medium having instructions that, if executed by a processing device, cause the processing device to perform operations comprising:
establishing an initial ranking of search results associated with a search query cluster comprising a first search query and a second search query;
identifying a first query identifier associated with the first search query received from a first end-user system;
identifying, based on the first query identifier, a set of model parameters associated with the first search query;
in response to receiving the first search query, generating, using a machine-learning model, one or more changes to configure a re-ranking of the search results based at least in part on the set of model parameters and the initial ranking of the search results;
associating the re-ranking of the search results with the search query cluster;
generating a first graphical user interface to display the re-ranking of the search results to the first end-user system;
executing, by the machine-learning model, a test using the one or more changes to generate a set of adjusted model parameters associated with the search query cluster;
identifying a second query identifier associated with the second search query received from a second end-user system; and
generating a second graphical user interface to display the re-ranking of the search results to the second end-user system.
14. The non-transitory computer readable storage medium of claim 13 , wherein the set of model parameters comprises an attractiveness score associated with each of the set of search results corresponding to the search query cluster.
15. The non-transitory computer readable storage medium of claim 14 , the operations further comprising swapping a position of a first search result and a second search result of the re-ranking of the search results to generate a further re-ranking of the search results.
16. The non-transitory computer readable storage medium of claim 13 , the operations further comprising, in response to receiving a subsequent submission of the first search query, causing a further re-ranking of the search results to be provided.
17. The non-transitory computer readable storage medium of claim 13 , wherein the set of adjusted model parameters is generated based at least in part on data points associated with search events collected over a period of time.
18. The non-transitory computer readable storage medium of claim 13 , the operations further comprising generating subsequent re-rankings of the search results based on the set of adjusted model parameters.