Granted Patent B2

US 9,940,372 · App. 14/720,867 · Granted Apr 10, 2018

Abstract

Claims (73)

1. A method comprising:

while a user is entering text to formulate a search query, and prior to receiving an indication that the user has completed formulation of the search query, performing the steps of:

determining a plurality of completed queries that include the text the user has already entered for the search query;

determining whether to proactively provide, to the user, search results for a particular completed query of the plurality of completed queries;

wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on at least one of:

whether the particular completed query qualifies as a dominant completed query;

a richness metric determined for the search results for the particular completed query; or

satisfaction metric determined for the search results for the particular completed query; and

responsive to determining to proactively provide search results for the particular completed query, causing presentation of the search results for the particular completed query to the user prior to receiving the indication that the user has completed formulation of the search query;

wherein the method is performed by one or more computing devices.

2. The method of claim 1 wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on whether the particular completed query qualifies as a dominant completed query.

3. The method of claim 2 further comprising:

determining a probable-intent score for the particular completed query;

wherein the probable-intent score indicates likelihood that the user intends to enter the particular completed query;

performing a comparison between the probable-intent score and a dominant-intent-threshold; and

determining whether the particular completed query qualifies as a dominant completed query based on the comparison between the probable-intent score and the dominant-intent-threshold.

4. The method of claim 3 further comprising determining the dominant-intent-threshold based, at least in part, on how many characters of text the user has entered.

5. The method of claim 1 wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the richness metric determined for the search results for the particular completed query.

6. The method of claim 5 wherein the richness metric is based, at least in part, on which types of content are contained in the search results for the particular completed query.

7. The method of claim 1 wherein:

the method further comprises monitoring how users that have previously submitted the particular completed query have interacted with the search results for the particular completed query;

based on how users that have previously submitted the particular completed query have interacted with the search results for the particular completed query, determining the satisfaction metric associated with the search results for the particular completed query; and

determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the satisfaction metric determined for the search results for the particular completed query.

8. The method of claim 1 wherein:

the method further comprises:

for each training query from a set of training queries, feeding a plurality of features relating to the training query into a machine learning tool to train the machine learning tool;

after training the machine learning tool, producing particular output by feeding particular features relating to the particular completed query to the machine learning tool; and

the step of determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the particular output.

9. The method of claim 1 wherein:

the step of determining a plurality of completed queries comprises:

identifying a first set of completed queries, wherein the first set of completed queries are completed queries that include the text;

determining a second set of completed queries, wherein the second set of completed queries are completed queries, from the first set, that satisfy certain criteria;

determining a third set of completed queries, wherein the third set of completed queries are completed queries, from the second set, that are most popular; and

the particular completed query is a most popular query from the third set of completed queries.

10. The method of claim 9 wherein the certain criteria includes that the queries were sent by users that match one or more characteristics of the user that is formulating the search query.

11. The method of claim 9 wherein the certain criteria includes that the queries were sent at a time that matches a time at which the user that is formulating the search query.

12. The method of claim 9 wherein the certain criteria includes that the queries were sent from locations that match location information related to the user that is formulating the search query.

13. One or more non-transitory storage media storing instructions which, when executed by one or more computing devices, cause performance of a method comprising:

while a user is entering text to formulate a search query, and prior to receiving an indication that the user has completed formulation of the search query, performing the steps of:

determining a plurality of completed queries that include the text the user has already entered for the search query;

determining whether to proactively provide, to the user, search results for a particular completed query of the plurality of completed queries;

wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on at least one of:

whether the particular completed query qualifies as a dominant completed query;

a richness metric determined for the search results for the particular completed query; or

satisfaction metric determined for the search results for the particular completed query; and

responsive to determining to proactively provide search results for the particular completed query, causing presentation of the search results for the particular completed query to the user prior to receiving the indication that the user has completed formulation of the search query.

14. The one or more non-transitory storage media of claim 13 wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on whether the particular completed query qualifies as a dominant completed query.

15. The one or more non-transitory storage media of claim 14 wherein the method further comprises:

determining a probable-intent score for the particular completed query;

wherein the probable-intent score indicates likelihood that the user intends to enter the particular completed query;

performing a comparison between the probable-intent score and a dominant-intent-threshold; and

determining whether the particular completed query qualifies as a dominant completed query based on the comparison between the probable-intent score and the dominant-intent-threshold.

16. The one or more non-transitory storage media of claim 15 wherein the method further comprises determining the dominant-intent-threshold based, at least in part, on how many characters of text the user has entered.

17. The one or more non-transitory storage media of claim 13 wherein determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the richness metric determined for the search results for the particular completed query.

18. The one or more non-transitory storage media of claim 17 wherein the richness metric is based, at least in part, on which types of content are contained in the search results for the particular completed query.

19. The one or more non-transitory storage media of claim 13 wherein:

the method further comprises monitoring how users that have previously submitted the particular completed query have interacted with the search results for the particular completed query;

based on how users that have previously submitted the particular completed query have interacted with the search results for the particular completed query, determining the satisfaction metric associated with the search results for the particular completed query; and

determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the satisfaction metric determined for the search results for the particular completed query.

20. The one or more non-transitory storage media of claim 13 wherein:

the method further comprises:

for each training query from a set of training queries, feeding a plurality of features relating to the training query into a machine learning tool to train the machine learning tool;

after training the machine learning tool, producing particular output by feeding particular features relating to the particular completed query to the machine learning tool; and

the step of determining whether to proactively provide, to the user, search results for the particular completed query is based, at least in part, on the particular output.

21. The one or more non-transitory storage media of claim 13 wherein:

the step of determining a plurality of completed queries comprises:

identifying a first set of completed queries, wherein the first set of completed queries are completed queries that include the text;

determining a second set of completed queries, wherein the second set of completed queries are completed queries, from the first set, that satisfy certain criteria;

determining a third set of completed queries, wherein the third set of completed queries are completed queries, from the second set, that are most popular; and

the particular completed query is a most popular query from the third set of completed queries.

22. The one or more non-transitory storage media of claim 21 wherein the certain criteria includes that the queries were sent by users that match one or more characteristics of the user that is formulating the search query.

23. The one or more non-transitory storage media of claim 21 wherein the certain criteria includes that the queries were sent at a time that matches a time at which the user that is formulating the search query.

24. The one or more non-transitory storage media of claim 21 wherein the certain criteria includes that the queries were sent from locations that match location information related to the user that is formulating the search query.

Assignments (6)