Method and system to answer a natural-language question

Abstract

Providing methods and systems to quickly and accurately respond to a natural-language question. The responses to the question can depend on additional information about the user asking the question, and the subject matter of the question the user asked. For example, the system knows that the user understands French, and can supply French answers to the user. Such additional information can improve on relevancy of the responses to the question. More than one responses can be provided to the user to allow the user to pick the more appropriate one. One embodiment uses a computer with a database having many phrases and question formats. The computer identifies phrases in the question based on at least one grammatical rules and phrases in the database. Then the computer links the phrases to categories based on at least one semantic rule, the subject matter of the question, and information about the user, such as previous questions asked by the user. The computer then selects at least two question formats based on at least the scores. After the question formats are selected, the system allows the user to pick at least one of the question formats so as to have an answer to the question generated.

We claim:  
     
         1 . A method for answering a natural-language question from a user by a first computing engine having access to a database with a plurality of phrases and question formats, the method comprising the steps of: 
 identifying at least one phrase in the question, based on phrases in the database and at least one grammatical rule;    linking the identified phrase to at least a category based on at least one semantic rule, and a piece of information about the user, other than the fact that the user asked the question; and    identifying at least two question formats in the database based on at least the category;    so that the user is allowed to pick at least one of the question formats for answering the question.    
     
     
         2 . A method as recited in  claim 1  wherein the method further comprises the step of providing a score to the identified phrase, with the score related to the category.  
     
     
         3 . A method as recited in  claim 2  wherein: 
 the step of identifying at least one phrase includes the step of identifying at least two phrases in the question, with the two phrases having at least one common word, based on phrases in the database and at least one grammatical rule;  
 the step of providing includes the step of providing a score to each of the identified phrases; and  
 the step of identifying at least two question formats includes the step of identifying at least two question formats in the database based on at least the scores.  
 
     
     
         4 . A method as recited in  claim 1  wherein the step of identifying at least one phrase depends on a hashing function.  
     
     
         5 . A method as recited in  claim 1  wherein the method further comprises the step of ignoring from the question a word that does not have a corresponding phrase in the database.  
     
     
         6 . A method as recited in  claim 1  wherein: 
 the step of identifying at least one phrase includes the step of identifying in the question every phrase that has a corresponding phrase in the database; and  
 the method further comprises the step of identifying a plurality of question structures based on the identified phrases.  
 
     
     
         7 . A method as recited in  claim 1  further comprising the step of selecting a plurality of question formats in the database based on the information about the user for identifying the at least two question formats.  
     
     
         8 . A method as recited in  claim 1  wherein the step of linking is also based on the subject matter of the question the user is asking.  
     
     
         9 . A method as recited in claim I wherein based on the question format picked, a URL is identified.  
     
     
         10 . A method as recited in  claim 1  wherein: 
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
 
     
     
         11 . A method as recited in  claim 1  wherein the piece of information about the user is related to the user's response to an inquiry from the computing engine, wherein the inquiry is for improving the accuracy of the answer to the question.  
     
     
         12 . A method as recited in  claim 11  wherein the piece of information about the user is also related to the user's response to another inquiry from the computing engine, presented to the user after the user has responded to the earlier inquiry.  
     
     
         13 . A method as recited in  claim 11  wherein the inquiry is on the subject matter of the question the user is asking.  
     
     
         14 . A method as recited in  claim 11  wherein the inquiry is related to an interest of the user.  
     
     
         15 . A method as recited in  claim 1  wherein the piece of information about the user is related to a question previously asked by the user.  
     
     
         16 . A method as recited in  claim 15  wherein the question previously asked by the user is stored in cookies.  
     
     
         17 . A method as recited in  claim 15  wherein: 
 a question format is written in HTML, with hidden variables;  
 the question previously asked by the user is stored in the hidden variables.  
 
     
     
         18 . A method as recited in  claim 1  wherein the piece of information about the user was not directly entered by the user.  
     
     
         19 . A method as recited in  claim 18  wherein: 
 the question was entered into a second computing engine, which is connected to the first computing engine through a network; and  
 the piece of information depends on the IP address of the second computing engine.  
 
     
     
         20 . A method as recited in  claim 2  wherein: 
 the question includes a plurality of phrases;  
 at least two of the phrases belong to two different categories;  
 the two different categories can be classified under a higher-level category; and  
 a phrase linked to the higher-level category has a higher score than phrases linked to the two different categories.  
 
     
     
         21 . A method as recited in  claim 1  wherein before the step of identifying at least one phrase in the question, the method further comprises the step of regularizing at least one phrase in the question.  
     
     
         22 . A method as recited in  claim 1  wherein: 
 the computing engine is for answering questions on a subject; and  
 a plurality of phrases and a plurality of question formats are for answering questions on the subject.  
 
     
     
         23 . A method as recited in  claim 1  wherein the category linked to the phrase is changed to another category in view of another piece of information about the user.  
     
     
         24 . A method for answering a natural-language question from a user by a computing engine having access to a database with a plurality of phrases and question formats, the method comprising the steps of: 
 identifying at least one phrase in the question, based on phrases in the database and at least one grammatical rule; and    identifying at least two question formats in the database based on the identified phrase and at least one semantic rule;    wherein: 
 the user is allowed to pick at least one of the question formats for answering the question;  
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
   
     
     
         25 . A method for answering a natural-language question from a user by a first computing engine, the method comprising the steps of: 
 transmitting, through a network, the question to a second computing engine having access to a database with a plurality of phrases and question formats, the second computing engine being configured to: 
 identify at least one phrase in the question, based on phrases in the database and at least one grammatical rule;  
 link the identified phrase to at least a category based on at least one semantic rule, and a piece of information about the user, other than the fact that the user asked the question; and  
 identify at least two question formats in the database based on at least the category;  
   so that the user is allowed to pick at least one of the question formats for answering the question.    
     
     
         26 . A method for answering a natural-language question from a user by a first computing engine, the method comprising the steps of: 
 transmitting, through a network, the question to a second computing engine having access to a database with a plurality of phrases and question formats, the second computing engine being configured to: 
 identify at least one phrase in the question, based on phrases in the database and at least one grammatical rule; and  
 identify at least two question formats in the database based on the identified phrase and at least one semantic rule;  
 wherein: 
 the user is allowed to pick at least one of the question formats for answering the question;  
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
 
   
     
     
         27 . An apparatus for answering a natural-language question from a user, the apparatus comprising: 
 a phrase identifier configured to identify at least one phrase in the question, based on phrases in a database and at least one grammatical rule;    a linker configured to link the identified phrase to at least a category based on at least one semantic rule, and a piece of information about the user, other than the fact that the user asked the question; and    a format identifier configured to identify at least two question formats in the database based on at least the category;    wherein the user is allowed to pick at least one of the question formats for answering the question.    
     
     
         28 . An apparatus for answering a natural-language question from a user, the apparatus comprising: 
 a phrase identifier configured to identify at least one phrase in the question, based on is phrases in a database and at least one grammatical rule; and    a format identifier configured to identify at least two question formats in the database based on the identified phrase and at least one semantic rule;    wherein: 
 the user is allowed to pick at least one of the question formats for answering the question;  
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
   
     
     
         29 . An apparatus for answering a natural-language question from a user comprising: 
 a transmitter configured to transmit, through a network, the question to a computing engine having access to a database with a plurality of phrases and question formats, the computing engine being configured to: 
 identify at least one phrase in the question, based on phrases in the database and at least one grammatical rule; and  
 link the identified phrase to at least a category based on at least one semantic rule, and a piece of information about the user, other than the fact that the user asked the question; and  
 identify at least two question formats in the database based on at least the category;  
   so that the user is allowed to pick at least one of the question formats for answering the question.    
     
     
         30 . An apparatus for answering a natural-language question from a user comprising: 
 a transmitter configured to transmit, through a network, the question to a computing engine having access to a database with a plurality of phrases and question formats, the computing engine being configured to: 
 identify at least one phrase in the question, based on phrases in the database and at least one grammatical rule; and  
 identify at least two question formats in the database based on the identified phrase and at least one semantic rule;  
 wherein: 
 the user is allowed to pick at least one of the question formats for answering the question;  
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
 
   
     
     
         31 . A computer readable media containing computer program code that is useful for answering a natural language question from a user, said code when executed by a first computer, having access to a database with a plurality of phrases and question formats, causing the first computer to perform a method comprising the steps of: 
 identifying at least one phrase in the question, based on phrases in the database and at least one grammatical rule;    linking the identified phrase to at least a category based on at least one semantic rule, and a piece of information about the user, other than the fact that the user asked the question; and    identifying at least two question formats in the database based on at least the category;    so that the user is allowed to pick at least one of the question formats for answering the question.    
     
     
         32 . A computer readable media containing computer program code that is useful for answering a natural language question from a user, said code when executed by a first computer, having access to a database with a plurality of phrases and question formats, causing the first computer to perform a method comprising the steps of: 
 identifying at least one phrase in the question, based on phrases in the database and at least one grammatical rule; and    identifying at least two question formats in the database based on the identified phrase and at least one semantic rule;    wherein: 
 the user is allowed to pick at least one of the question formats for answering the question;  
 one phrase in the question is being translated into a language different from the language of the question; and  
 an answer to the question in the different language is identified.  
   
     
     
         33 . A method for answering a natural-language question from a user by a first computing engine having access to a database with a plurality of phrases and question formats, the method comprising the steps of: 
 identifying at least one phrase in the question, based on phrases in the database and at least one grammatical rule;    linking the identified phrase to at least a category based on at least one semantic rule, and the subject matter of the question the user is asking; and    identifying at least two question formats in the database based on at least the category;    so that the user is allowed to pick at least one of the question formats for answering the question;    wherein the method is applicable to more than one subject matter based on the same database.