Providing result-based query suggestions
In general, one aspect of the subject matter described can be embodied in a method that includes, obtaining a plurality of search results responsive to an initial search query, the search results including a first search result that identifies a first resource; determining, using a document-to-query-to-document model, that the first resource is relevant to a first suggested query different from the initial search query; generating a presentation of the search results responsive to the initial search query; and providing the presentation of the search results in response to the initial search query. Each search result in the presentation includes a link to a respective resource, wherein the first search result in the presentation includes a link that, upon a selection by a user, can cause the first suggested query to be submitted to a search engine.
1. A method for generating and using a document-to-query-to-document model, the method comprising:
generating a set of document-to-query models, each document-to-query model associating a document of the document-to-query model with one or more queries for which the document has a relevance measure that satisfies a threshold; and
for each particular query of each particular document-to-query model:
identifying, in a set of query-to-document models, one or more query-to-document models generated for the particular query, wherein each query-to-document model generated for the particular query associates the particular query with one or more documents for which the relevance measure for the particular query satisfies the threshold; and
attaching the identified one or more query-to-document models to the particular query of each document-to-query model that includes the particular query as one of the one or more queries of the document-to-query model, thereby generating the document-to-query-to-document model that associates documents with related documents through queries including associating, through the particular query, the document of the particular document-to-query-model to the one or more documents associated with the particular query; and
using the document-to-query-to-document model to provide at least one of (i) search results or (ii) suggested queries in response to received search queries that are in the document-to-query-to-document model.
2. The method of claim 1 , further comprising:
identifying a plurality of search results in response to a received search query;
determining, using the document-to-query-to-document model, that a first resource identified by a first search result of the plurality of search results is relevant to a first query associated with the first resource in the document-to-query-to-document model; and
providing a presentation of the plurality of search results and the first query in response to the received search query.
3. The method of claim 1 , wherein generating the set of document-to-query models comprises inverting a set of query-to-document models, each query-to-document model in the set of query-to-document models associating a given query with one or more documents for which the relevance measure for the given query satisfies the threshold.
4. The method of claim 1 , wherein the relevance measure for a given document and a given query is based on a frequency with which users interact with search results that identify the given document when the search results are provided in response to receiving the given query.
5. The method of claim 1 , wherein identifying one or more query-to-document models comprises:
identifying the set of query-to-document models;
identifying a number of documents associated with a given query by a given query-to-document model;
determining that the number of documents is less than a threshold number of documents; and
filtering the given query-to-document model from the set of query-to-document models in response to determining that the number of documents is less than the threshold number of documents.
6. The method of claim 1 , further comprising:
for a given document-to-query model:
identifying two queries associated with the document of the given document-to-query model that have at least a threshold similarity to one another; and
removing one of the two queries from the given document-to-query model in response to identifying the two queries.
7. The method of claim 1 , wherein identifying one or more query-to-document models comprises:
identifying the set of query-to-document models;
identifying a number of documents associated with a given query by a given query-to-document model;
determining that the number of documents is greater than a threshold number of documents; and
removing one or more documents from the given query-to-document model in response to determining that the number of documents is greater than the threshold number of documents.
8. A system comprising:
one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
generating a set of document-to-query models, each document-to-query model associating a document of the document-to-query model with one or more queries for which the document has a relevance measure that satisfies a threshold; and
for each particular query of each particular document-to-query model:
identifying, in a set of query-to-document models, one or more query-to-document models generated for the particular query, wherein each query-to-document model generated for the particular query associates the particular query with one or more documents for which the relevance measure for the particular query satisfies the threshold; and
attaching the identified one or more query-to-document models to the particular query of each document-to-query model that includes the particular query as one of the one or more queries of the document-to-query model, thereby generating the document-to-query-to-document model that associates documents with related documents through queries including associating, through the particular query, the document of the particular document-to-query-model to the one or more documents associated with the particular query; and
using the document-to-query-to-document model to provide at least one of (i) search results or (ii) suggested queries in response to received search queries that are in the document-to-query-to-document model.
9. The system of claim 8 , wherein the operations further comprise:
identifying a plurality of search results in response to a received search query;
determining, using the document-to-query-to-document model, that a first resource identified by a first search result of the plurality of search results is relevant to a first query associated with the first resource in the document-to-query-to-document model; and
providing a presentation of the plurality of search results and the first query in response to the received search query.
10. The system of claim 8 , wherein generating the set of document-to-query models comprises inverting a set of query-to-document models, each query-to-document model in the set of query-to-document models associating a given query with one or more documents for which the relevance measure for the given query satisfies the threshold.
11. The system of claim 8 , wherein the relevance measure for a given document and a given query is based on a frequency with which users interact with search results that identify the given document when the search results are provided in response to receiving the given query.
12. The system of claim 8 , wherein identifying one or more query-to-document models comprises:
identifying the set of query-to-document models;
identifying a number of documents associated with a given query by a given query-to-document model;
determining that the number of documents is less than a threshold number of documents; and
filtering the given query-to-document model from the set of query-to-document models in response to determining that the number of documents is less than the threshold number of documents.
13. The system of claim 8 , wherein the operations further comprise:
for a given document-to-query model:
identifying two queries associated with the document of the given document-to-query model that have at least a threshold similarity to one another; and
removing one of the two queries from the given document-to-query model in response to identifying the two queries.
14. The system of claim 8 , wherein identifying one or more query-to-document models comprises:
identifying the set of query-to-document models;
identifying a number of documents associated with a given query by a given query-to-document model;
determining that the number of documents is greater than a threshold number of documents; and
removing one or more documents from the given query-to-document model in response to determining that the number of documents is greater than the threshold number of documents.
15. A computer program product, encoded on one or more non-transitory computer storage media, comprising instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:
generating a set of document-to-query models, each document-to-query model associating a document of the document-to-query model with one or more queries for which the document has a relevance measure that satisfies a threshold; and
for each particular query of each particular document-to-query model:
identifying, in a set of query-to-document models, one or more query-to-document models generated for the particular query, wherein each query-to-document model generated for the particular query associates the particular query with one or more documents for which the relevance measure for the particular query satisfies the threshold; and
attaching the identified one or more query-to-document models to the particular query of each document-to-query model that includes the particular query as one of the one or more queries of the document-to-query model, thereby generating the document-to-query-to-document model that associates documents with related documents through queries including associating, through the particular query, the document of the particular document-to-query-model to the one or more documents associated with the particular query; and
using the document-to-query-to-document model to provide at least one of (i) search results or (ii) suggested queries in response to received search queries that are in the document-to-query-to-document model.
16. The computer program product of claim 15 , wherein the operations further comprise:
identifying a plurality of search results in response to a received search query;
determining, using the document-to-query-to-document model, that a first resource identified by a first search result of the plurality of search results is relevant to a first query associated with the first resource in the document-to-query-to-document model; and
providing a presentation of the plurality of search results and the first query in response to the received search query.
17. The computer program product of claim 15 , wherein generating the set of document-to-query models comprises inverting a set of query-to-document models, each query-to-document model in the set of query-to-document models associating a given query with one or more documents for which the relevance measure for the given query satisfies the threshold.
18. The computer program product of claim 15 , wherein the relevance measure for a given document and a given query is based on a frequency with which users interact with search results that identify the given document when the search results are provided in response to receiving the given query.
19. The computer program product of claim 15 , wherein identifying one or more query-to-document models comprises:
identifying the set of query-to-document models;
identifying a number of documents associated with a given query by a given query-to-document model;
determining that the number of documents is less than a threshold number of documents; and
filtering the given query-to-document model from the set of query-to-document models in response to determining that the number of documents is less than the threshold number of documents.
20. The computer program product of claim 15 , wherein the operations further comprise:
for a given document-to-query model:
identifying two queries associated with the document of the given document-to-query model that have at least a threshold similarity to one another; and
removing one of the two queries from the given document-to-query model in response to identifying the two queries.