Associating features with entities, such as categories of web page documents, and/or weighting such features

Abstract

Features that may be used to represent relevance information (e.g., properties, characteristics, etc.) of an entity, such as a document or concept for example, may be associated with the document by accepting an identifier that identifies a document; obtaining search query information (and/or other serving parameter information) related to the document using the document identifier, determining features using the obtained query information (and/or other serving parameter information), and associating the features determined with the document. Weights of such features may be similarly determined. The weights may be determined using scores. The scores may be a function of one or more of whether the document was selected, a user dwell time on a selected document, whether or not a conversion occurred with respect to the document, etc. The document may be a Web page. The features may be n-grams. The relevance information of the document may be used to target the serving of advertisements with the document.

1 . A computer-implemented method comprising: 
 a) obtaining serving information related to a document;    b) determining features using the obtained serving information; and    c) associating the features determined with the document.    
   
   
       2 . The computer-implemented method of  claim 1  further comprising: 
 d) determining whether or not to serve an ad with the document using the features associated with the document.    
   
   
       3 . The computer-implemented method of  claim 1  wherein the serving information related to the document includes information from at least one past query that caused the rendering of information of the document on a search results list.  
   
   
       4 . The computer-implemented method of  claim 3  wherein the serving information related to the document includes whether or not the rendered information of the document was selected.  
   
   
       5 . The computer-implemented method of  claim 3  wherein the serving information related to the document includes a time that the user dwelled on the document after selecting the rendered information of the document.  
   
   
       6 . The computer-implemented method of  claim 3  wherein the document is a Web page.  
   
   
       7 . The computer-implemented method of  claim 6  wherein the serving information related to the document includes information from at least one past query that caused the rendering of information of the document on a search results list.  
   
   
       8 . The computer-implemented method of  claim 6  wherein the serving information related to the document includes whether or not the rendered information of the document was selected.  
   
   
       9 . The computer-implemented method of  claim 6  wherein the serving information related to the document includes a time that the user dwelled on the document after selecting the rendered information of the document.  
   
   
       10 . The computer-implemented method of  claim 1  further comprising: 
 e) obtaining user action information related to the document using the document identifier;    f) determining scores using the user action information; and    g) assigning weights to the features using the scores determined.    
   
   
       11 . The computer-implemented method of  claim 10  wherein each of the weights is a monotonic function of an associated one of the scores.  
   
   
       12 . The computer-implemented method of  claim 11  wherein the user action is a dwell time after a selection, and wherein the score is higher for a longer dwell time than for a shorter dwell time.  
   
   
       13 . The computer-implemented method of  claim 11  wherein the user action is selection, and wherein the score is higher for a selection than for a non-selection.  
   
   
       14 . The computer-implemented method of  claim 11  wherein the user action is conversion, and wherein the score is higher for a conversion than for a non-conversion.  
   
   
       15 . The computer-implemented method of  claim 1  further comprising: 
 e) determining scores using the serving information ; and    f) assigning weights to the features using the scores determined.    
   
   
       16 . The computer-implemented method of  claim 15  wherein the score for a feature is determined using a frequency of the feature in the serving information .  
   
   
       17 . The computer-implemented method of  claim 15  wherein the score for a feature is determined using an inverse frequency of the feature in serving information for a collection of documents.  
   
   
       18 . The computer-implemented method of  claim 1  further comprising: 
 e) obtaining user action information related to the document using the document identifier;    f) determining scores using both the serving information and the user action information; and    g) assigning weights to the features using the scores determined.    
   
   
       19 . The computer-implemented method of  claim 18  wherein each of the weights is a monotonic function of an associated one of the scores.  
   
   
       20 . The computer-implemented method of  claim 1  further comprising: 
 e) determining scores using at least one of (A) the serving information and (B) user action information related to the document; and    f) filtering the features using the scores determined.    
   
   
       21 . The computer-implemented method of  claim 20  further comprising: 
 g) assigning weights to the features using the scores determined.    
   
   
       22 . The computer-implemented method of  claim 1  wherein at least one of the features is an n-gram.  
   
   
       23 . The computer-implemented method of  claim 1  wherein at least one of the features is a keyword.  
   
   
       24 . The computer-implemented method of  claim 1  wherein at least one of the features is a concept.  
   
   
       25 . The computer-implemented method of  claim 1  wherein the serving information related to the document is obtained using an accepted document identifier.  
   
   
       26 . The computer-implemented method of  claim 25  wherein the document identifier is a universal resource locator.  
   
   
       27 . A computer-implemented method comprising: 
 a) accepting a feature-to-entity association;    b) using the feature-to-entity association to generate one or more results for presentation to a user;    c) tracking user behavior with respect to the results; and    d) updating a score associated with the feature-to-entity association using the tracked user behavior.    
   
   
       28 . The computer-implemented method of  claim 27  wherein the feature-to-entity association is a keyword-to-category association.  
   
   
       29 . The computer-implemented method of  claim 28  wherein the one or more results generated include one or more category listings provided on a document.  
   
   
       30 . The computer-implemented method of  claim 27  wherein the feature-to-entity association is a category-to-ad association.  
   
   
       31 . The computer-implemented method of  claim 30  wherein the one or more results generated include one or more category targeted ads provided on a document.  
   
   
       32 . The computer-implemented method of  claim 27  wherein the user behavior includes whether or not a user selects a result.  
   
   
       33 . The computer-implemented method of  claim 27  wherein the user behavior includes whether or not a user converts on a result.  
   
   
       34 . Apparatus comprising: 
 a) means for obtaining serving information related to a document;    b) means for determining features using the obtained serving information;    and    c) means for associating the features determined with the document.    
   
   
       35 . Apparatus comprising: 
 a) means for accepting a feature-to-entity association;    b) means for using the feature-to-entity association to generate one or more results for presentation to a user;    c) means for tracking user behavior with respect to the results; and    d) means for updating a score associated with the feature-to-entity association using the tracked user behavior.