Systems and methods for identifying claims in electronic text

Abstract

A system, method, and computer program for identifying claims associated with electronic text are provided. In an approach, electronic text is accessed. Linguistic content associated with the electronic text is identified. A linguistic structure is generated based on the linguistic content identified. The linguistic structure is compared to a claim template. A claim is identified based on the comparison.

What is claimed is: 
     
         1 . A data processing method comprising:
 receiving a corpus of documents containing one or more product claim annotations;   identifying, based on the one or more product claim annotations in the corpus of documents, one or more template linguistic structures likely to indicate presence of a product claim;   accessing electronic text;   identifying linguistic content associated with the electronic text, wherein the linguistic content includes a plurality of linguistic features;   generating a linguistic structure based on the linguistic content identified, wherein the linguistic structure identifies at least a relationship between the plurality of linguistic features;   identifying a particular product claim within the electronic text based on comparing the linguistic structure to the one or more template linguistic structures;   wherein the method is performed by one or more computing devices.   
     
     
         2 . The method of  claim 1 , wherein identifying the one or more template linguistic structures comprises using one or more machine learning techniques including: maximum entropy, support vector machine, neural network, nearest neighbor, hidden Markov model, conditional random fields, or maximum entropy Markov model. 
     
     
         3 . The method of  claim 1 , wherein the one or more product claim annotations identify at least one or more boundaries of a claim within the corpus of documents and a type categorization for the claim. 
     
     
         4 . The method of  claim 1 , wherein the one or more template linguistic structures specify one or more features including: lexical entities, grammatical relations, semantic meanings, or argumentative structure. 
     
     
         5 . The method of  claim 1 , wherein identifying the claim within the electronic text involves tagging the claim within the electronic text with one or more annotations identifying the linguistic structure. 
     
     
         6 . The method of  claim 5 , wherein the one or more product claim annotations specify features including: product name, product type, product benefit, object of the product benefit, or product user category. 
     
     
         7 . The method of  claim 1 , further comprising:
 determining whether the particular product claim is misleading based on one or more factors including: absence of risk information, absence of required supporting references, non- compliance with a set of rules, presence of false benefits, incorrect side effect information, omissions of material facts, obfuscated risk disclosure, amount or type of evidence presented for the particular product claim, presence of anecdotal evidence, references to government agency evaluations, or references to historical or traditional use of a product;   in response to determining that the particular product claim is misleading, storing data that flags the particular product claim as misleading.   
     
     
         8 . The method of  claim 1 , further comprising: generating, based on the particular product claim, one or more product recommendations for a user. 
     
     
         9 . The method of  claim 8 , wherein the one or more product recommendations include one or more of: a list of suitable products or services for the user, an advertisement to the user identifying one or more products, or data specifying one or more sellers which sell the one or more products. 
     
     
         10 . The method of  claim 8 , wherein generating the one or more product recommendations is based on input supplied by the user. 
     
     
         11 . The method of  claim 1 , further comprising: generating, based on the particular product claim, one or more of: a product description, a product review, or a comparison between one or more different products. 
     
     
         12 . A computer-readable storage medium storing one or more instructions which, when executed by one or more processors, cause the one or more processors to perform:
 receiving a corpus of documents containing one or more product claim annotations;   identifying, based on the one or more product claim annotations in the corpus of documents, one or more template linguistic structures likely to indicate presence of a product claim;   accessing electronic text;   identifying linguistic content associated with the electronic text, wherein the linguistic content includes a plurality of linguistic features;   generating a linguistic structure based on the linguistic content identified, wherein the linguistic structure identifies at least a relationship between the plurality of linguistic features;   identifying a particular product claim within the electronic text based on comparing the linguistic structure to the one or more template linguistic structures.   
     
     
         13 . The computer-readable storage medium of  claim 12 , wherein the instructions for identifying the one or more template linguistic structures comprise instructions which when executed by the one or more processors cause using one or more machine learning techniques including: maximum entropy, support vector machine, neural network, nearest neighbor, hidden Markov model, conditional random fields, or maximum entropy Markov model 
     
     
         14 . The computer-readable storage medium of  claim 12 , wherein the one or more product claim annotations identify at least one or more boundaries of a claim within the corpus of documents and a type categorization for the claim. 
     
     
         15 . The computer-readable storage medium of  claim 12 , wherein the one or more template linguistic structures specify one or more features including: lexical entities, grammatical relations, semantic meanings, or argumentative structure. 
     
     
         16 . The computer-readable storage medium of  claim 12 , wherein the instructions which when executed cause identifying the claim within the electronic text comprise instructions which when executed by the one or more processors cause tagging the claim within the electronic text with one or more annotations identifying the linguistic structure. 
     
     
         17 . The computer-readable storage medium of  claim 16 , the one or more product claim annotations specify features including: product name, product type, product benefit, object of the product benefit, or product user category. 
     
     
         18 . The computer-readable storage medium of  claim 12 , comprising instructions which when executed by the one or more processors cause performing:
 determining whether the particular product claim is misleading based on one or more factors including: absence of risk information, absence of required supporting references, non- compliance with a set of rules, presence of false benefits, incorrect side effect information, omissions of material facts, obfuscated risk disclosure, amount or type of evidence presented for the particular product claim, presence of anecdotal evidence, references to government agency evaluations, or references to historical or traditional use of a product;   in response to determining that the particular product claim is misleading, storing data that flags the particular product claim as misleading.   
     
     
         19 . The computer-readable storage medium of  claim 12 , comprising instructions which when executed by the one or more processors cause performing generating, based on the particular product claim, one or more product recommendations for a user. 
     
     
         20 . The computer-readable storage medium of  claim 19 , wherein the one or more product recommendations include one or more of: a list of suitable products or services for the user, an advertisement to the user identifying one or more products, or data specifying one or more sellers which sell the one or more products. 
     
     
         21 . The computer-readable storage medium of  claim 19 , comprising instructions which when executed by the one or more processors cause generating the one or more product recommendations is based on input supplied by the user. 
     
     
         22 . The computer-readable medium of  claim 12 , comprising instructions which when executed by the one or more processors cause generating, based on the particular product claim, one or more of: a product description, a product review, or a comparison between one or more different products. 
     
     
         23 . A data processing method comprising:
 accessing electronic text;   identifying linguistic content associated with the electronic text, wherein the linguistic content includes a plurality of linguistic features;   generating a linguistic structure based on the linguistic content identified, wherein the linguistic structure identifies at least a relationship between the plurality of linguistic features;   identifying a particular product claim within the electronic text based on comparing the linguistic structure to a claim template;   generating, based on the particular product claim, one or more product recommendations for a user;   wherein the method is performed by one or more computing devices.   
     
     
         24 . The method of  claim 23 , wherein the one or more product recommendations include one or more of: a list of suitable products or services for the user, an advertisement to the user identifying one or more products, data specifying one or more sellers which sell the one or more products, a product description, a product review, or a comparison between one or more different products. 
     
     
         25 . The method of  claim 23 , wherein generating the one or more product recommendations is based on input supplied by the user. 
     
     
         26 . A computer-readable storage medium storing one or more instructions which, when executed by one or more processors, cause the one or more processors to perform:
 accessing electronic text;   identifying linguistic content associated with the electronic text, wherein the linguistic content includes a plurality of linguistic features;   generating a linguistic structure based on the linguistic content identified, wherein the linguistic structure identifies at least a relationship between the plurality of linguistic features;   identifying a particular product claim within the electronic text based on comparing the linguistic structure to a claim template;   generating, based on the particular product claim, one or more product recommendations for a user;   wherein the method is performed by one or more computing devices.   
     
     
         27 . The computer-readable medium of  claim 26 , wherein the one or more product recommendations include one or more of: a list of suitable products or services for the user, an advertisement to the user identifying one or more products, data specifying one or more sellers which sell the one or more products, a product description, a product review, or a comparison between one or more different products. 
     
     
         28 . The computer-readable medium of  claim 26 , wherein generating the one or more product recommendations is based on input supplied by the user.