Robust anchoring of annotations to content

Abstract

Annotations are robustly anchored to content. The robust anchoring allows annotations to be associated with the content independently of the format used to store the content. The annotations can be re-anchored to the content after the content has been modified, including modifications to the format used to store the content.

1 - 34 . (canceled)  
   
   
       35 . A method of determining a portion of modified content to which an annotation corresponds, the method comprising: 
 receiving annotation anchoring information, wherein the annotation anchoring information describes a portion of the content that the annotation corresponded to prior to the portion being modified;    identifying keywords in the modified content;    determining, for a region of the content that potentially corresponds to the annotation, a score for the region based at least in part on the location of the keywords in the content;    repeating the determining for a plurality of identified keywords in the content; and    selecting the region having the best score as the portion of content to which the annotation corresponds.    
   
   
       36 . A method as recited in  claim 35 , further comprising selecting the region having the best score as the portion of content to which the annotation corresponds if the best score exceeds a lower bound, and otherwise selecting no region as the portion of content to which the annotation corresponds.  
   
   
       37 . A method as recited in  claim 35 , further comprising selecting the region having the best score as the portion of content to which the annotation corresponds without user input if the best score exceeds a threshold value, and otherwise obtaining user input prior to selecting the region having the best score as the portion of content to which the annotation corresponds.  
   
   
       38 . A method as recited in  claim 37 , wherein the user input comprises a user request for a best guess as to the region to which the annotation corresponds.  
   
   
       39 . A method as recited in  claim 35 , further comprising selecting the region having the best score as the portion of content to which the annotation corresponds and stopping the repeating if the best score exceeds a threshold value.  
   
   
       40 . A method as recited in  claim 35 , wherein the content and the modified content are stored in two different formats.  
   
   
       41 . A method as recited in  claim 35 , wherein selecting the region having the best score comprises selecting the region having the highest score.  
   
   
       42 . A method as recited in  claim 35 , wherein determining the score for the region comprises: 
 identifying each keyword in the annotation anchoring information;    selecting one of the keywords in the annotation anchoring information;    locating the selected keyword in the modified content;    locating one or more additional keywords that are both identified in the annotation anchoring information and that are within a particular distance in the modified content of the selected keyword; and    increasing the score for each of the one or more additional keywords.    
   
   
       43 . A method as recited in  claim 42 , wherein the particular distance comprises twice the length of the portion of the content that the annotation corresponding to prior to the portion being modified.  
   
   
       44 . A method as recited in  claim 35 , wherein determining the score for the region comprises: 
 selecting a keyword from the annotation anchoring information;    identifying, from the annotation anchoring information, one or more beginning point features for the portion of the content;    locating the selected keyword in the modified content;    locating a point of the modified content that precedes the selected keyword and that matches the one or more beginning point features; and    determining whether to increase the score for the region including the selected keyword and the located point based at least in part on the distance between the located point and the selected keyword.    
   
   
       45 . A method as recited in  claim 44 , wherein determining whether to increase the score for the region comprises determining to increase the score for the region if the distance from the located point to the selected keyword is not greater than a threshold amount.  
   
   
       46 . A method as recited in  claim 44 , wherein the located point comprises an intermediate point.  
   
   
       47 . A method as recited in  claim 35 , wherein determining the score for the region comprises: 
 selecting a keyword from the annotation anchoring information;    identifying, from the annotation anchoring information, one or more ending point features for the portion of the content;    locating the selected keyword in the modified content;    locating a point of the modified content that succeeds the selected keyword and that matches the one or more ending point features; and    determining whether to increase the score for the region including the selected keyword and the located point based at least in part on the distance between the selected keyword and the located point.    
   
   
       48 . A method as recited in  claim 47 , wherein determining whether to increase the score for the region comprises determining to increase the score for the region if the distance from the selected keyword to the located point is not greater than a threshold amount.  
   
   
       49 . A method as recited in  claim 47 , wherein the located point comprises an intermediate point.  
   
   
       50 - 56 . (canceled)  
   
   
       57 . One or more computer readable media having stored thereon a plurality of instructions that, when executed by one or more processors, causes the one or more processors to perform acts comprising: 
 receiving an annotation anchor that describes a region of original content to which an annotation corresponds;    receiving a modified version of the original content;    identifying one or more key parts in the annotation anchor;    identifying, based at least in part on the one or more key parts, one or more regions of the modified version to which the annotation potentially corresponds;    generating a score for the one or more regions; and    selecting, as the region of the modified version of the original content to which the annotation corresponds, the identified region having the best score.    
   
   
       58 . One or more computer readable media as recited in  claim 57 , wherein the region of original content has been altered in the modified version.  
   
   
       59 . One or more computer readable media as recited in  claim 57 , wherein the selecting comprises: 
 checking whether the best score exceeds a threshold value;    automatically selecting the identified region having the best score if the best score exceeds the threshold value; and    otherwise, querying a user for input as to whether the identified region is the region of the modified version to which the annotation corresponds.    
   
   
       60 . One or more computer readable media as recited in  claim 57 , wherein the plurality of instructions further cause the one or more processors to perform acts comprising: 
 generating a new annotation anchor based on captured features of the identified region; and    using the new annotation anchor as an anchor that describes the region of the modified version of the original content to which the annotation corresponds.    
   
   
       61 . One or more computer readable media as recited in  claim 57 , wherein the key parts are keywords, and wherein the plurality of instructions further cause the one or more processors to perform acts comprising: 
 identifying one or more synonyms for at least one of the one or more keywords; and    altering the score for a region if at least one of the one or more synonyms is included in the region.    
   
   
       62 - 74 . (canceled)