Enhancing performance of structured lookups using set operations

Abstract

A system and computer program product configured to train a system to be able to provide answers to questions that do not have a direct relationship established in a database.

1 .- 12 . (canceled) 
     
     
         13 . A system comprising a processor configured to:
 receive a training question and a training answer set to the training question;   determine a training question structure of the training question;   determine a training subject in the training question;   determine a plurality of paths in a database that links the training subject to a plurality of answer subsets, wherein an answer subset comprises at least one answer in the training answer set to the training question;   perform a relation scoring analysis on the plurality of answer subsets to determine the answer subset or a combination of the answer subsets that provides a highest accuracy score, wherein the relation scoring analysis is configured to perform an operation on the combination of the answer subsets based on a precision and a recall of the answer subsets in comparison to the training answer set to the training question; and   store a subset of the plurality of paths corresponding to the answer subset or the combination of the answer subsets having the highest accuracy score, the operation performed on the combination of the answer subsets having the highest accuracy score, and the training question structure of the training question for use in providing answers to questions.   
     
     
         14 . The system of  claim 13 , wherein the processor is further configured to:
 receive a non-training question;   determine a subject in the non-training question;   determine a question structure of the non-training question;   determine whether the question structure of the non-training question matches the training question structure of the training question;   retrieve, from a database, the answer subset or the combination of the answer subsets corresponding to the subject in the non-training question by applying the subset of the plurality of paths corresponding to the answer subset or the combination of the answer subsets having the highest accuracy score to the subject of the non-training question in response to determining that the question structure of the non-training question matches the training question structure of the training question;   return the answer subset as the answer to the non-training question in response to the answer subset being retrieved;   apply the operation on the combination of the answer subsets to generate a modified answer subset in response to the combination of the answer subsets being retrieved; and   return the modified answer subset as the answer to the non-training question.   
     
     
         15 . The system of  claim 13 , wherein the operation is a union operation and wherein the relation scoring analysis further comprises:
 performing the union operation on the combination of the answer subsets for the answer subsets that have the recall of less than one; and   returning the combination of the answer subsets in which the union operation of the combination of the answer subsets provides the highest accuracy score.   
     
     
         16 . The system of  claim 13 , wherein the operation is an intersection operation and wherein the relation scoring analysis further comprises:
 performing the intersection operation on the combination of the answer subsets for the answer subsets that have the recall equal to one; and   returning the combination of the answer subsets in which the intersection operation of the combination of the answer subsets provides the highest accuracy score.   
     
     
         17 . A computer program product for training a system to enable the system to provide answers to questions that do not have a direct relationship established in a database, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
 receive a training question and a training answer set to the training question;   determine a training question structure of the training question;   determine a training subject in the training question;   determine a plurality of paths in a database that links the training subject to a plurality of answer subsets, wherein an answer subset comprises at least one answer in the training answer set to the training question;   perform a relation scoring analysis on the plurality of answer subsets to determine the answer subset or a combination of the answer subsets that provides a highest accuracy score, wherein the relation scoring analysis is configured to perform an operation on the combination of the answer subsets based on a precision and a recall of the answer subsets in comparison to the training answer set to the training question; and   store a subset of the plurality of paths corresponding to the answer subset or the combination of the answer subsets having the highest accuracy score, the operation performed on the combination of the answer subsets having the highest accuracy score, and the training question structure of the training question for use in providing answers to questions.   
     
     
         18 . The computer program product of  claim 17 , further comprising program instructions executable by the processor to cause the processor to:
 receive a non-training question;   determine a subject in the non-training question;   determine a question structure of the non-training question;   determine whether the question structure of the non-training question matches the training question structure of the training question;   retrieve, from a database, the answer subset or the combination of the answer subsets corresponding to the subject in the non-training question by applying the subset of the plurality of paths corresponding to the answer subset or the combination of the answer subsets having the highest accuracy score to the subject of the non-training question in response to determining that the question structure of the non-training question matches the training question structure of the training question;   return the answer subset as the answer to the non-training question in response to the answer subset being retrieved;   apply the operation on the combination of the answer subsets to generate a modified answer subset in response to the combination of the answer subsets being retrieved; and   return the modified answer subset as the answer to the non-training question.   
     
     
         19 . The computer program product of  claim 17 , further comprising program instructions executable by the processor to cause the processor to:
 perform the union operation on the combination of the answer subsets for the answer subsets that have the recall of less than one; and   return the combination of the answer subsets in which the union operation of the combination of the answer subsets provides the highest accuracy score.   
     
     
         20 . The computer program product of  claim 17 , further comprising program instructions executable by the processor to cause the processor to:
 perform the intersection operation on the combination of the answer subsets for the answer subsets that have the recall equal to one; and   return the combination of the answer subsets in which the intersection operation of the combination of the answer subsets provides the highest accuracy score.