Multi-term query subsumption for document classification

Abstract

In general, embodiments of the present invention provide systems, methods and computer readable media for generating an optimal classifying query set for categorizing and/or labeling textual data based on a query subsumption calculus to determine, given two queries, whether one of the queries subsumes another. In one aspect, a method includes generating a group of determining queries based on analyzing text within a document; receiving a group of classifying queries; and, for each determining query within the group of determining queries, determining whether at least one of the classifying queries is subsumed by the determining query; and updating the group of classifying queries in an instance in which the classifying query is subsumed by the determining query.

1 - 27 . (canceled) 
     
     
         28 . An apparatus, comprising one or more processors and one or more storage devices storing instructions that are operable, when executed by the one or more processors, to cause the one or more processors to:
 generate a classifying queries subset associated with document text within a document based at least in part on respective text spans associated with a length of a respective list of features for a respective classifying query of the classifying queries subset that respectively represents a particular category for the document text;   determine a document categorization score based at least in part on a normalized sum of respective performance metrics for respective classifying queries of the classifying queries subset; and   correlate the document with the particular category represented by the classifying queries subset in an instance in which the document categorization score is greater than a categorization threshold value.   
     
     
         29 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 modify the classifying queries subset in response to a determination that a classifying query of the classifying queries subset is subsumed by a query associated with at least one term identified within the document text.   
     
     
         30 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 modify the classifying queries subset in response to a determination that a classifying query of the classifying queries subset and a query associated with the document text share at least two common terms.   
     
     
         31 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 add a query associated with at least one term identified within the document text to the classifying queries subset based at least in part on a comparison between the respective text spans and a text span for the query.   
     
     
         32 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 remove a classifying query from the classifying queries subset based at least in part on a comparison between a text span for the classifying query and one or more other text spans for one or more other classifying queries of the classifying queries subset.   
     
     
         33 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 generate the classifying queries subset associated with the document text within the document based at least in part on an ordering of features in the respective list of features for the respective classifying query of the classifying queries.   
     
     
         34 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 generate the classifying queries subset associated with the document text within the document based at least in part on a difference between at least a first term and a second term within the respective classifying query of the classifying queries.   
     
     
         35 . The apparatus of  claim 28 , wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to:
 generate the respective list of features for the respective classifying query of the classifying queries subset based at least in part on a classification model associated with supervised learning.   
     
     
         36 . A computer-implemented method, comprising:
 generating, by a computing device comprising a processor, a classifying queries subset associated with document text within a document based at least in part on respective text spans associated with a length of a respective list of features for a respective classifying query of the classifying queries subset that respectively represents a particular category for the document text;   determining, by the computing device, a document categorization score based at least in part on a normalized sum of respective performance metrics for respective classifying queries of the classifying queries subset; and   correlating, by the computing device, the document with the particular category represented by the classifying queries subset in an instance in which the document categorization score is greater than a categorization threshold value.   
     
     
         37 . The computer-implemented method of  claim 36 , further comprising:
 modifying, by the computing device, the classifying queries subset in response to a determination that a classifying query of the classifying queries subset is subsumed by a query associated with at least one term identified within the document text.   
     
     
         38 . The computer-implemented method of  claim 36 , further comprising:
 modifying, by the computing device, the classifying queries subset in response to a determination that a classifying query of the classifying queries subset and a query associated with the document text share at least two common terms.   
     
     
         39 . The computer-implemented method of  claim 36 , further comprising:
 adding, by the computing device, a query associated with at least one term identified within the document text to the classifying queries subset based at least in part on a comparison between the respective text spans and a text span for the query.   
     
     
         40 . The computer-implemented method of  claim 36 , further comprising:
 removing, by the computing device, a classifying query from the classifying queries subset based at least in part on a comparison between a text span for the classifying query and one or more other text spans for one or more other classifying queries of the classifying queries subset.   
     
     
         41 . The computer-implemented method of  claim 36 , wherein the generating the classifying queries subset comprises generating the classifying queries subset based at least in part on an ordering of features in the respective list of features for the respective classifying query of the classifying queries. 
     
     
         42 . The computer-implemented method of  claim 36 , wherein the generating the classifying queries subset comprises generating the classifying queries subset based at least in part on a difference between at least a first term and a second term within the respective classifying query of the classifying queries. 
     
     
         43 . The computer-implemented method of  claim 36 , further comprising:
 generating, by the computing device, the respective list of features for the respective classifying query of the classifying queries subset based at least in part on a classification model associated with supervised learning.   
     
     
         44 . A computer program product, stored on a computer readable medium, comprising instructions that when executed by one or more computers cause the one or more computers to:
 generate a classifying queries subset associated with document text within a document based at least in part on respective text spans associated with a length of a respective list of features for a respective classifying query of the classifying queries subset that respectively represents a particular category for the document text;   determine a document categorization score based at least in part on a normalized sum of respective performance metrics for respective classifying queries of the classifying queries subset; and   correlate the document with the particular category represented by the classifying queries subset in an instance in which the document categorization score is greater than a categorization threshold value.   
     
     
         45 . The computer program product of  claim 44 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 generate the classifying queries subset associated with the document text within the document based at least in part on an ordering of features in the respective list of features for the respective classifying query of the classifying queries.   
     
     
         46 . The computer program product of  claim 44 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 generate the classifying queries subset associated with the document text within the document based at least in part on a difference between at least a first term and a second term within the respective classifying query of the classifying queries.   
     
     
         47 . The computer program product of  claim 44 , further comprising instructions that when executed by the one or more computers cause the one or more computers to:
 generate the respective list of features for the respective classifying query of the classifying queries subset based at least in part on a classification model associated with supervised learning.