Query entity-experience classification

Abstract

The technology described herein makes improved use of limited screen space on a search results page by determining whether to present a question-and-answer experience and/or an entity details experience. This determination effects the amount of information presented and the format in which it is presented. In general, the question-and-answer experience provides less information and is more targeted to a question and query terms other than the entity. In contrast, the entity details experience provides more information about the entity that is not tailored to the query beyond the entity being included in the query. In one aspect, the determination of whether to show a question-and-answer experience and/or an entity details experience is based, at least in part, on an entity-details intent classification score (“intent classification score”) generated by a machine classification system. The classification score may be processed in combination with other criteria to make a final determination.

What is claimed is: 
     
         1 . A method for providing relevant search result content in an experience format that is optimized to a query, comprising:
 receiving, at a search engine, a query from a user, the query including an entity;   determining an entity-details intent classification score for the query using a machine classifier;   identifying a type for a question-and-answer experience that would be presented by the search engine in response to the query;   determining that an entity-details intent is associated with the query when the entity-details intent classification score is above a threshold score;   determining that the type for the question-and-answer experience satisfies a type criteria; and   causing the question-and-answer experience to be output to the user, rather than an entity-details experience, because the question-and-answer experience satisfies the type criteria.   
     
     
         2 . The method of  claim 1 , wherein the type criteria is the question-and-answer experience matching a plurality of question-and-answer types. 
     
     
         3 . The method of  claim 2 , wherein the plurality of question-and-answer types include a healthcare response and a list response. 
     
     
         4 . The method of  claim 1 , wherein the type criteria is the question-and-answer experience being generated from a source on a curated source list. 
     
     
         5 . The method of  claim 4 , wherein the curated source list includes educational reference documents. 
     
     
         6 . The method of  claim 1 , wherein the method further comprises training the machine classifier using labeled edge case training queries that are identified using a preliminary classifier trained using a small set of labeled training queries, wherein the small set comprises less than 1/10 a number of total training queries used to train the machine classifier. 
     
     
         7 . The method of  claim 1 , wherein the method further comprises determining that the query matches a specific query within a hotfix list, wherein the hotfix list includes one or more specific queries that should not be assigned an entity-experience intent. 
     
     
         8 . A method for providing relevant search result content in an experience format that is optimized to a query comprising:
 receiving, at a search engine, a query from a user, the query including an entity;   determining an entity-details intent classification score for the query using a machine classifier;   identifying a type for a question-and-answer experience that would be presented by the search engine in response to the query;   determining that an entity-details intent is not associated with the query when the entity-details intent classification score is below a threshold score;   determining that the type for the question-and-answer experience satisfies a type criteria; and   causing the question-and-answer experience to be output to the user, rather than an entity-details experience, because the entity-details intent is not associated with the query.   
     
     
         9 . The method of  claim 8 , wherein the type criteria is the entity not exactly matching a title of the entity-details experience. 
     
     
         10 . The method of  claim 8 , further comprising identifying information about the question-and-answer experience that would be presented by the search engine in response to the query, wherein the type criteria is a similarity score between content in the question-and-answer experience having greater than a threshold similarity to content in the entity-details experience. 
     
     
         8 . The method of  claim 8 , further comprising:
 identifying information about the question-and-answer experience that would be presented by the search engine in response to the query,   wherein the type criteria comprises the question-and-answer experience matching a plurality of question-and-answer types on a list.   
     
     
         9 . The method of  claim 8 , further comprising:
 identifying information about the question-and-answer experience that would be presented by the search engine in response to the query,   wherein the type criteria comprises the question-and-answer experience being generated from a source on a curated source list.   
     
     
         10 . The method of  claim 8 , wherein the machine classifier is a support vector machine. 
     
     
         11 . The method of  claim 12 , wherein the question-and-answer experience includes a subset of content about the entity that is selected from a full set of content about the entity based on terms in the query different from terms that identify the entity. 
     
     
         12 . The method of  claim 8 , wherein the method further comprises determining that the query matches a specific query within a hotfix list, wherein the hotfix list includes one or more specific queries that should not be associated with an entity-experience intent. 
     
     
         13 . One or more computer storage media that, when executed by a computing device, causes the computing device to providing a relevant search result content in an experience format that is optimized to a query, the method comprising:
 receiving, at a search engine, a query from a user, the query including an entity;   determining an entity-details intent classification score for the query using a machine classifier;   identifying information about a question-and-answer experience that would be presented by the search engine in response to the query;   identifying information about an entity-details experience that would be presented by the search engine in response to the query;   determining that an entity-details intent is not associated with the query using a plurality of rules that take the entity-details intent classification score, the information about the question-and-answer experience, and the information about the entity-details experience as input; and   outputting the question-and-answer experience to the user.   
     
     
         14 . The media of  claim 13 , wherein one of the plurality of rules comprises the entity-details intent classification score being above a threshold score and the question-and-answer experience matching a plurality of question-and-answer types on a curated source list. 
     
     
         15 . The media of  claim 16 , wherein one of the plurality of rules comprises the entity-details intent classification score being below a threshold score. 
     
     
         16 . The media of  claim 13 , wherein one of the plurality of rules comprises the entity not exactly matching a title of the entity-details experience. 
     
     
         17 . The media of  claim 16 , wherein one of the plurality of rules comprises a similarity score between content in the question-and-answer experience not having greater than a threshold similarity to content in the entity-details experience.