Selecting content using query-independent scores of query segments

Abstract

Methods, systems, and apparatus include computer programs encoded on a computer-readable storage medium, including a method for responding to queries. A first user query is received. The first user query is processed including identifying one or more segments in the first user query, a segment representing a word or a phrase. A stand-alone score is determined for each segment of the first user query, wherein the stand-alone score is an indication of a likelihood that the segment represents a stand-alone query and that the segment represents a main topic of the first user query. A historical log of queries is processed to determine query-independent scores for segments that are included in queries represented by the log. The final query-independent scores are used to determine the stand-alone score for each segment of the first query.

What is claimed is: 
     
         1 . (canceled) 
     
     
         2 . A computer-implemented method performed by one or more processors comprising:
 determining a score for each segment of a query, wherein determining the score includes:
 determining a query-independent score for the segment, wherein the query-independent score specifies a likelihood that the segment represents a stand-alone query; 
 determining a query-dependent for the segment, wherein the query-dependent specifies a likelihood that the segment is dependent on other segments in the query; and 
 determining the score for the segment based on the query-dependent and query-independent scores. 
   
     
     
         3 . The computer-implemented method of  claim 2 , wherein each segment comprises a plurality of consecutive words in the query. 
     
     
         4 . The computer-implemented method of  claim 2 , further comprising:
 determining, for queries in a historical log of queries, query-independent scores for segments of the queries;   determining, for the queries in the historical log of queries, query-dependent scores for the segments, including normalizing the query-dependent scores for a given query; and   adjusting a score for each segment of a query using a mathematical function that incorporates each of the query-independent scores for the segments and normalized query-dependent scores for the segments.   
     
     
         5 . The computer-implemented method of  claim 4 , wherein the mathematical function comprises an average function or a sum function. 
     
     
         6 . The computer-implemented method of  claim 2 , further comprising:
 executing a process including:
 identifying candidate content items from an inventory to serve in response to the query; 
 identifying one or more keywords associated with a given one of the candidate content items; 
 determining one or more query-independent scores for each of the one or more keywords; and 
 determining a self-sufficiency score for a given candidate content item based on the one or more query-independent scores for the one or more keywords associated with a given one of the candidate content items; 
   repeating the process for other ones of the candidate content items from the inventory;   comparing a self-sufficiency score of the query to self-sufficiency scores of the ones of the candidate content items to locate a match; and   providing a matching content item responsive to the query.   
     
     
         7 . The computer-implemented method of  claim 6 , further comprising for all matching content items located, conducting an auction to determine which matching content item to use when providing the matching content item. 
     
     
         8 . The computer-implemented method of  claim 7 , further comprising:
 computing a sum of scores for all keywords associated with a content item in the inventory;   comparing the sum of scores to a first threshold; and   disqualifying a content item for inclusion in the auction when the first threshold is not met.   
     
     
         9 . The computer-implemented method of  claim 6 , further comprising:
 computing a sum of scores for all segments in the query; and   not using the self-sufficiency score to select a content item when the sum of scores is below a second threshold.   
     
     
         10 . A non-transitory computer-readable medium storing instructions, that when executed, cause one or more processors to perform operations comprising:
 determining a score for each segment of a query, wherein determining the score includes:
 determining a query-independent score for the segment, wherein the query-independent score specifies a likelihood that the segment represents a stand-alone query; 
 determining a query-dependent for the segment, wherein the query-dependent specifies a likelihood that the segment is dependent on other segments in the query; and 
 determining the score for the segment based on the query-dependent and query-independent scores. 
   
     
     
         11 . The non-transitory computer-readable medium of  claim 10 , further comprising:
 determining, for queries in a historical log of queries, query-independent scores for segments of the queries;   determining, for the queries in the historical log of queries, query-dependent scores for the segments, including normalizing the query-dependent scores for a given query; and   adjusting a score for each segment of a query using a mathematical function that incorporates each of the query-independent scores for the segments and normalized query-dependent scores for the segments.   
     
     
         12 . The non-transitory computer-readable medium of  claim 10 , further comprising:
 executing a process including:
 identifying candidate content items from an inventory to serve in response to the query; 
 identifying one or more keywords associated with a given one of the candidate content items; 
 determining one or more query-independent scores for each of the one or more keywords; and 
 determining a self-sufficiency score for a given candidate content item based on the one or more query-independent scores for the one or more keywords associated with a given one of the candidate content items; 
   repeating the process for other ones of the candidate content items from the inventory;   comparing a self-sufficiency score of the query to self-sufficiency scores of the ones of the candidate content items to locate a match; and   providing a matching content item responsive to the query.   
     
     
         13 . The non-transitory computer-readable medium of  claim 12 , further comprising for all matching content items located, conducting an auction to determine which matching content item to use when providing the matching content item. 
     
     
         14 . The non-transitory computer-readable medium of  claim 13 , further comprising:
 computing a sum of scores for all keywords associated with a content item in the inventory;   comparing the sum of scores to a first threshold; and   disqualifying a content item for inclusion in the auction when the first threshold is not met.   
     
     
         15 . The non-transitory computer-readable medium of  claim 10 , further comprising:
 computing a sum of scores for all segments in the query; and   not using the self-sufficiency score to select a content item when the sum of scores is below a second threshold.   
     
     
         16 . A system comprising:
 one or more processors; and   one or more memory elements including instructions that, when executed, cause the one or more processors to perform operations comprising:
 determining a score for each segment of a query, wherein determining the score includes:
 determining a query-independent score for the segment, wherein the query-independent score specifies a likelihood that the segment represents a stand-alone query; 
 determining a query-dependent for the segment, wherein the query-dependent specifies a likelihood that the segment is dependent on other segments in the query; and 
 determining the score for the segment based on the query-dependent and query-independent scores. 
 
   
     
     
         17 . The system of  claim 16 , further comprising:
 determining, for queries in a historical log of queries, query-independent scores for segments of the queries;   determining, for the queries in the historical log of queries, query-dependent scores for the segments, including normalizing the query-dependent scores for a given query; and   adjusting a score for each segment of a query using a mathematical function that incorporates each of the query-independent scores for the segments and normalized query-dependent scores for the segments.   
     
     
         18 . The system of  claim 16 , further comprising:
 executing a process including:
 identifying candidate content items from an inventory to serve in response to the query; 
 identifying one or more keywords associated with a given one of the candidate content items; 
 determining one or more query-independent scores for each of the one or more keywords; and 
 determining a self-sufficiency score for a given candidate content item based on the one or more query-independent scores for the one or more keywords associated with a given one of the candidate content items; 
   repeating the process for other ones of the candidate content items from the inventory;   comparing a self-sufficiency score of the query to self-sufficiency scores of the ones of the candidate content items to locate a match; and   providing a matching content item responsive to the query.   
     
     
         19 . The system of  claim 18 , further comprising for all matching content items located, conducting an auction to determine which matching content item to use when providing the matching content item. 
     
     
         20 . The system of  claim 19 , further comprising:
 computing a sum of scores for all keywords associated with a content item in the inventory;   comparing the sum of scores to a first threshold; and   disqualifying a content item for inclusion in the auction when the first threshold is not met.   
     
     
         21 . The system of  claim 16 , further comprising:
 computing a sum of scores for all segments in the query; and   not using the self-sufficiency score to select a content item when the sum of scores is below a second threshold.