Query refinements using search data
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining query refinements using search data. In one aspect, a method includes receiving a first query and a second query each comprising one or more n-grams for a user session, determining a first set of query refinements for the first query, determining a second set of query refinements from the first set of query refinements, each query refinement in the second set of query refinements including at least one n-gram that is similar to an n-gram from the first query and at least on n-gram that is similar to an n-gram from the second query, scoring each query refinement in the second set of query refinements, selecting a third query from a group consisting of the second set of query refinements and the second query, and providing the third query as input to a search operation.
1. A system comprising one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
receiving, from a user device, a first query comprising one or more first n-grams for a user session;
receiving, from the user device, a second query comprising one or more second n-grams for the user session, the second query being different from the first query, comprising one or more second n-grams, not including all of the one or more first n-grams, and subsequent to the first query in the user session without an intervening query;
determining, from historical data, two or more query refinements, each query refinement in the two or more query refinements including at least one n-gram that is similar to one of the one or more first n-grams and at least one n-gram that is similar to one of the one or more second n-grams;
scoring each query refinement in the two or more query refinements by assigning a higher score to a first query refinement from the two or more query refinements that was similar to a query refinement received from a device associated with a user account for the user device than a second query refinement from the two or more query refinements that is not similar to a query refinement received from a device associated with the user account for the user device, each score for a specific query refinement based on one or more query refinement n-grams that are similar to one of the one or more first n-grams or one of the one or more second n-grams, the query refinement n-grams included in the specific query refinement;
selecting, based on the scores, a highest scored query refinement from the two or more query refinements; and
providing, to the user device, first data for first search results responsive to the highest scored query refinement.
2. The system of claim 1 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement from the two or more query refinements that was similar to a query refinement received from the user device than the second query refinement from the two or more query refinements that is not similar to a query refinement received from the user device associated with the user account for the user device.
3. The system of claim 2 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement that was received from the user device than the second query refinement that was not received from the user device.
4. The system of claim 1 , the operations comprising:
determining, for at least one of the two or more query refinements, a likelihood that the query refinement will be received by a search engine sequentially after the search engine receives the first query, wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements using the likelihood that a query refinement will be received by a search engine sequentially after the search engine receives the first query.
5. The system of claim 1 , wherein the first query refinement is similar to the query refinement received from a device associated with the user account for the user device when i) a first n-gram from the first query refinement and a second n-gram from the query refinement are an exact match, synonyms, alternate spellings of a same n-gram, or n-grams from a same class, or ii) when one of the first n-gram or the second n-gram is an acronym for the other n-gram, an expansion for the other n-gram, an abbreviation for the other n-gram, or a different grammatical number n-gram for the other n-gram.
6. The system of claim 1 , wherein selecting the highest scored query refinement comprises:
determining whether a highest score for the highest scored query refinement satisfies a threshold score; and
in response to determining that the highest score for the highest scored query refinement satisfies the threshold score, selecting the highest scored query refinement from the two or more query refinements that has the highest score.
7. The system of claim 1 , wherein scoring each query refinement comprises scoring each query refinement using the historical data.
8. A non-transitory computer storage medium encoded with instructions that, when executed by one or more computers, cause the one or more computers to perform operations comprising:
receiving, from a user device, a first query comprising one or more first n-grams for a user session;
receiving, from the user device, a second query comprising one or more second n-grams for the user session, the second query being different from the first query, comprising one or more second n-grams, not including all of the one or more first n-grams, and subsequent to the first query in the user session without an intervening query;
determining, from historical data, two or more query refinements, each query refinement in the two or more query refinements including at least one n-gram that is similar to one of the one or more first n-grams and at least one n-gram that is similar to one of the one or more second n-grams;
scoring each query refinement in the two or more query refinements by assigning a higher score to a first query refinement from the two or more query refinements that was similar to a query refinement received from a device associated with a user account for the user device than a second query refinement from the two or more query refinements that is not similar to a query refinement received from a device associated with the user account for the user device, each score for a specific query refinement based on one or more query refinement n-grams that are similar to one of the one or more first n-grams or one of the one or more second n-grams, the query refinement n-grams included in the specific query refinement;
selecting, based on the scores, a highest scored query refinement from the two or more query refinements; and
providing, to the user device, first data for first search results responsive to the highest scored query refinement.
9. The computer storage medium of claim 8 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement from the two or more query refinements that was similar to a query refinement received from the user device than the second query refinement from the two or more query refinements that is not similar to a query refinement received from the user device associated with the user account for the user device.
10. The computer storage medium of claim 9 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement that was received from the user device than the second query refinement that was not received from the user device.
11. The computer storage medium of claim 8 , the operations comprising:
determining, for at least one of the two or more query refinements, a likelihood that the query refinement will be received by a search engine sequentially after the search engine receives the first query, wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements using the likelihood that a query refinement will be received by a search engine sequentially after the search engine receives the first query.
12. The computer storage medium of claim 8 , wherein the first query refinement is similar to the query refinement received from a device associated with the user account for the user device when i) a first n-gram from the first query refinement and a second n-gram from the query refinement are an exact match, synonyms, alternate spellings of a same n-gram, or n-grams from a same class, or ii) when one of the first n-gram or the second n-gram is an acronym for the other n-gram, an expansion for the other n-gram, an abbreviation for the other n-gram, or a different grammatical number n-gram for the other n-gram.
13. The computer storage medium of claim 8 , wherein selecting the highest scored query refinement comprises:
determining whether a highest score for the highest scored query refinement satisfies a threshold score; and
in response to determining that the highest score for the highest scored query refinement satisfies the threshold score, selecting the highest scored query refinement from the two or more query refinements that has the highest score.
14. The computer storage medium of claim 8 , wherein scoring each query refinement comprises scoring each query refinement using the historical data.
15. A computer-implemented method comprising:
receiving, from a user device, a first query comprising one or more first n-grams for a user session;
receiving, from the user device, a second query comprising one or more second n-grams for the user session, the second query being different from the first query, comprising one or more second n-grams, not including all of the one or more first n-grams, and subsequent to the first query in the user session without an intervening query;
determining, from historical data, two or more query refinements, each query refinement in the two or more query refinements including at least one n-gram that is similar to one of the one or more first n-grams and at least one n-gram that is similar to one of the one or more second n-grams;
scoring each query refinement in the two or more query refinements by assigning a higher score to a first query refinement from the two or more query refinements that was similar to a query refinement received from a device associated with a user account for the user device than a second query refinement from the two or more query refinements that is not similar to a query refinement received from a device associated with the user account for the user device, each score for a specific query refinement based on one or more query refinement n-grams that are similar to one of the one or more first n-grams or one of the one or more second n-grams, the query refinement n-grams included in the specific query refinement;
selecting, based on the scores, a highest scored query refinement from the two or more query refinements; and
providing, to the user device, first data for first search results responsive to the highest scored query refinement.
16. The method of claim 15 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement from the two or more query refinements that was similar to a query refinement received from the user device than the second query refinement from the two or more query refinements that is not similar to a query refinement received from the user device associated with the user account for the user device.
17. The method of claim 16 , wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements by assigning a higher score to the first query refinement that was received from the user device than the second query refinement that was not received from the user device.
18. The method of claim 15 , comprising:
determining, for at least one of the two or more query refinements, a likelihood that the query refinement will be received by a search engine sequentially after the search engine receives the first query, wherein scoring each query refinement in the two or more query refinements comprises scoring each query refinement in the two or more query refinements using the likelihood that a query refinement will be received by a search engine sequentially after the search engine receives the first query.
19. The method of claim 15 , wherein the first query refinement is similar to the query refinement received from a device associated with the user account for the user device when i) a first n-gram from the first query refinement and a second n-gram from the query refinement are an exact match, synonyms, alternate spellings of the same n-gram, or n-grams from the same class, or ii) when one of the first n-gram or the second n-gram is an acronym for the other n-gram, an expansion for the other n-gram, an abbreviation for the other n-gram, or a different grammatical number n-gram for the other n-gram.
20. The method of claim 15 , wherein the first query refinement is similar to the query refinement received from a device associated with the user account for the user device when i) a first n-gram from the first query refinement and a second n-gram from the query refinement are an exact match, synonyms, alternate spellings of a same n-gram, or n-grams from a same class, or ii) when one of the first n-gram or the second n-gram is an acronym for the other n-gram, an expansion for the other n-gram, an abbreviation for the other n-gram, or a different grammatical number n-gram for the other n-gram.