Back to results
US 7,984,004 B2Granted
Query suggestion generation
AssigneeMicrosoft Corporation
InventorsGalen Andrew, Sooho Park, Robert L. Rounthwaite, Silviu-Petru Cucerzan, Jamie Paul Buckley, Joanna Chan
Life of this patent
Estimating the term…
Abstract
Described herein is a system that facilitates assigning indications of usefulness to query suggestions. The system includes a query suggestion generator component that receives a query and generates a query suggestion based at least in part upon the received query. A model component outputs an indication of usefulness with respect to the query suggestion, wherein the model component is a machine-learned model of user behavior with respect to query suggestions.
Claims18
A system that facilitates outputting indications of usefulness of query suggestions, the system comprising:
- a processor;
- and
- a memory that comprises a plurality of components that are executable by the processor, the plurality of components comprising:
- a query suggestion generator component that receives a query and generates a query suggestion based at least in part upon the received query;
- a model component that outputs an indication of usefulness with respect to the query suggestion, wherein the model component is a logistic regression model that has been trained to model user interaction with query suggestions output by a search engine, wherein the logistic regression model has been trained based at least in part upon queries provided to the search engine by users, search results returned to the users subsequent to the search engine performing searches using the queries, search results selected by the users subsequent to the search results being returned to the users, query suggestions provided to the users responsive to the queries being provided to the search engine by the users, query suggestions selected by the users upon the users being provided with the query suggestions, and search results selected by the users subsequent to the users selecting the query suggestions;
- and
- a grouping component that groups query suggestions into a plurality of different groups, wherein the model component individually assigns an indication of usefulness to each of the groups of query suggestions, wherein the groups are assembled based at least in part upon an amount of overlap between search results that correspond to query suggestions.
4 more elements
A method that facilitates determining an indication of usefulness of a query suggestion with respect to a query, comprising the following computer-executable acts:
- receiving the query from a user;
- generating a plurality of query suggestions based at least in part upon the received query;
- causing a processor to create multiple groups of query suggestions from the plurality of generated query suggestions, wherein each of the multiple groups of query suggestions includes a plurality of query suggestions;
- using a logistic regression model that has been trained from data that is indicative of user interaction with respect to query suggestions to output an indication of usefulness for each group of queries in the multiple groups of query suggestions based at least in part upon the received query, wherein query suggestions in a group of query suggestions with a highest indication of usefulness assigned thereto are presented to the user, wherein the data that is indicative of user interaction with respect to query suggestions comprises queries provided to a search engine by users, search results returned to the users subsequent to the search engine performing searches using the queries, search results selected by the users subsequent to the search results being returned to the users, query suggestions provided to the users responsive to the queries being provided to the search engine by the users, query suggestions selected by the users upon the users being provided with the query suggestions, and search results selected by the users subsequent to the users selecting the query suggestions.
1 more element
A computer-readable memory comprising instructions that, when executed by a processor, cause the processor to perform acts comprising:
- receiving, at a search engine, a query issued by a user;
- responsive to receiving the query, locating a plurality of candidate query suggestions that are based at least in part upon the query;
- responsive to receiving the query, accessing a logistic regression model that has been trained to model user interaction with query suggestions output by the search engine, wherein the logistic regression model has been trained based at least in part upon queries provided to a search engine by users, search results returned to the users subsequent to the search engine performing searches using the queries, search results selected by the users subsequent to the search results being returned to the users, query suggestions provided to the users responsive to the queries being provided to the search engine by the users, query suggestions selected by the users upon the users being provided with the query suggestions, and search results selected by the users subsequent to the users selecting the query suggestions;
- providing each query suggestion in the plurality of candidate query suggestions to the logistic regression model, wherein the logistic regression model is configured to output an indication of usefulness for each of the query suggestions in the plurality of candidate query suggestions upon receipt of each of the plurality of candidate query suggestions;
- and
- presenting at least one query suggestion from amongst the plurality of candidate query suggestions to the user subsequent to receiving the query issued by the user, wherein the at least one query suggestion is selected for presentment based at least in part upon the indication of usefulness for the query suggestion.
3 more elements
Dependencies are inferred from the claim text.