Providing answers to questions using logical synthesis of candidate answers

• US 10,133,808 B2
• Filed: 05/02/2016
• Issued: 11/20/2018
• Est. Priority Date: 09/28/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Brown, Eric W., Chu-Carroll, Jennifer, Ferrucci, David A., Lally, Adam P., Murdock, James W., Prager, John M.
• Primary Examiner(s): Le, Hung
• Application Number:
  US15/144,328
• Publication Number:
  US 20160246875A1
• Time in Patent Office: 932 Days
• Field of Search: None
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3344 using natural language analysis

Abstract

A method, system and computer program product for generating answers to questions. In one embodiment, the method comprises receiving an input query, decomposing the input query into a plurality of different subqueries, and conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries. A ranking function is applied to each of the candidate answers to determine a ranking for each of these candidate answers; and for each of the subqueries, one of the candidate answers to the subquery is selected based on this ranking. A logical synthesis component is applied to synthesize a candidate answer for the input query from the selected the candidate answers to the subqueries. In one embodiment, the procedure applied by the logical synthesis component to synthesize the candidate answer for the input query is determined from the input query.

Claims

1. A method of generating an answer to a question, comprising:

receiving an input query;

using natural language processing in a question processing module for processing the input query to determine a plurality of sub-queries, wherein each of the subqueries has an associated answer including a specified data item of a defined type determined by said each sub-query, and the specified data items of the answers to the subqueries are used to determine an answer to the input query;

using the natural language processing in the question processing module for processing the input query to determine a procedure for using the answers to the subqueries to determine the answer to the input query;

conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries, wherein the at least one candidate answer to each of the sub-queries is of the defined type by said each sub-query;

for each of the subqueries, selecting one of the candidate answers to the subquery; and

using said procedure determined using the natural language processing in the question processing module for processing the input query, to combine the specified data items of the selected ones of the candidate answers to the sub-queries to determine a candidate answer to the input query.

2. The method according to claim 1, wherein the using said determined procedure includes applying a logical synthesis component for synthesizing the candidate answer for the input query from the selected candidate answers to the subqueries.

3. The method according to claim 1, wherein the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes determining said procedure when using the input query to determine the plurality of subqueries.

4. The method according to claim 3, wherein:

the using the input query to determine the plurality of subqueries includes using natural language processing to decompose the input query; and

the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes using said natural language processing to determine said procedure.

5. The method according to claim 1, wherein the selecting one of the candidate answers for each of the subqueries includes applying a ranking function to rank the candidate answers for each of the subqueries.

6. The method according to claim 1, wherein the using said determined procedure to combine the specified data items of the selected ones of the candidate answers to determine a candidate answer to the input query includes using a logical synthesis component to perform a mathematical operation on data in the selected candidate answers to synthesize the candidate answer for the input query.

8. The method according to claim 1, wherein using said determined procedure to combine the specified data items of the selected ones of the candidate answers to determine a candidate answer to the input query includes using a logical synthesis component to form a union of the selected candidate answers to synthesize the candidate answer for the input query.

9. The method according to claim 1, wherein the using said determined procedure to combine the specified data items of the selected ones of the candidate answers to determine a candidate answer to the input query includes using a logical synthesis component to combine text from the selected answers to synthesize the candidate answer for the input query.

10. The method according to claim 1, further comprising applying a candidate ranking function to the candidate answer for the input query to determine a ranking for the candidate answer for the input query.

7. A method of generating an answer to a question, comprising:

receiving an input query;

using the input query to determine a plurality of sub-queries, wherein each of the subqueries has an associated answer, and the answers to the subqueries are used to determine an answer to the input query;

conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries;

for each of the subqueries, selecting one of the candidate answers to the subquery;

using the input query to determine a procedure for using the selected ones of the candidate answers to determine a candidate answer to the input query; and

using said determined procedure to use the selected ones of the candidate answers to determine a candidate answer to the input query, and wherein;

the using said determined procedure to use the selected ones of the candidate answers to determine a candidate answer to the input query includes using a logical synthesis component to perform a mathematical operation on data in the selected candidate answers to synthesize the candidate answer for the input query; and

each of the selected candidate answers to the subqueries includes an associated date, and the using a logical synthesis component to perform a mathematical operation on data in the selected candidate answers to synthesize the candidate answer includes using the associated dates of the selected candidate answers to determine a time duration.

11. A system for generating answers to questions, comprising:

a computer device comprising at least one distinct software module, each distinct software module being embodied on a tangible computer-readable medium;

a memory; and

at least one processor coupled to the memory and operative for;

receiving an input query;

using natural language processing in a question processing module for processing the input query to determine a plurality of sub-queries;

wherein each of the subqueries has an associated answer including a specified data item of a defined type determined by said each sub-query, and the specified data items of the answers to the subqueries are used to determine an answer to the input query;

using the natural language processing in the question processing module for processing the input query to determine a procedure for using the answers to the subqueries to determine the answer to the input query;

conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries, wherein the at least one candidate answer to each of the sub-queries is of the defined type determined by said each sub-query;

for each of the subqueries, selecting one of the candidate answers to the subquery; and

using said procedure determined using the natural language processing in the question processing module for processing the input query, to combine the specified data items of the selected ones of the candidate answers to the sub-queries to determine a candidate answer to the input query.

12. The system according to claim 11, wherein the using said determined procedure includes applying a logical synthesis component for synthesizing the candidate answer for the input query from the selected candidate answers to the subqueries .

13. The system according to claim 12, wherein the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes determining said procedure when using the input query to determine the plurality of subqueries.

14. The system according to claim 11, wherein the using the input query to determine the plurality of subqueries includes using natural language processing to decompose the input query;

and

the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes using said natural language processing to determine said procedure.

15. The system according to claim 11, wherein the selecting one of the candidate answers for each of the subqueries include applying a ranking function to rank the candidate answers for each of the subqueries.

16. An article of manufacture, comprising:

at least one tangible computer readable medium having computer readable program code logic to generate answers to questions, the computer readable program code logic, when executing, performing the following;

receiving an input query;

using natural language processing in a question processing module for processing the input query to determine a plurality of sub-queries, wherein each of the subqueries has an associated answer including a specified data item of a defined type determined by said each sub-query, and the specified data items of the answers to the subqueries are used to determine an answer to the input query;

using the natural language processing in the question processing module for processing the input query to determine a procedure for using the answers to the subqueries to determine the answer to the input query;

conducting a search in one or more data sources to identify at least one candidate answer to each of the subqueries, wherein the at least one candidate answer to each of the sub-queries is of the defined type determined by said each sub-query;

for each of the subqueries, selecting one of the candidate answers to the subquery; and

using said procedure determined using the natural language processing in the question processing module for processing the input query, to combine the specified data items of the selected ones of the candidate answers to the sub-queries to determine a candidate answer to the input query.

17. The article of manufacture according to claim 16, wherein the using said determined procedure includes applying a logical synthesis component for synthesizing the candidate answer for the input query from the selected candidate answers to the subqueries.

18. The article of manufacture according to claim 16, wherein the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes determining said procedure when using the input query to determine the plurality of subqueries.

19. The article of manufacture according to claim 16, wherein:

the using the input query to determine the plurality of subqueries includes using natural language processing to decompose the input query; and

the determining the procedure for combining the specified data items of the selected ones of the candidate answers includes using said natural language processing to determine said procedure.

20. The article of manufacture according to claim 16, wherein the selecting one of the candidate answers for each of the subqueries include applying a ranking function to rank the candidate answers for each of the subqueries.