Identification and evaluation of lexical answer type conditions in a question to generate correct answers

• US 10,176,228 B2
• Filed: 12/10/2014
• Issued: 01/08/2019
• Est. Priority Date: 12/10/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bishop, Timothy A., Boxwell, Stephen A., Brumfield, Benjamin L., Desai, Nirav P., Vernier, Stanley J.
• Primary Examiner(s): Burke, Jeff A
• Assistant Examiner(s): Vu, Thong
• Application Number:
  US14/565,672
• Publication Number:
  US 20160170989A1
• Time in Patent Office: 1,490 Days
• Field of Search: 707723, 707738, 707780, 707776, 707748, 707713, 707739, 705 272
• CPC Class Codes: G06F 16/23 Updating
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/24573 using data annotations, e.g. user-defined metadata
• CPC Class Codes: G06F 16/24575 using context
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/248 Presentation of query results

Abstract

An approach is provided in which a knowledge manager receives a question and selects candidate answers based upon a lexical answer type included in the question. The knowledge manager evaluates the candidate answers against conditions detected in the question and generates condition fulfillment scores accordingly. In turn, the knowledge manager ranks the candidate answers based upon the corresponding condition fulfillment scores.

Claims

1. A method implemented by an information handling system that includes a memory and a processor, the method comprising:

receiving a question that includes a logical answer type;

identifying, by a processor, one or more terms in the question that impose one or more requirements of the question, wherein the one or more terms are different from the logical answer type;

annotating the question, by the processor, by adding one or more condition annotations to the question that correspond to the one or more requirements, wherein at least one of the one or more condition annotations are selected from the group consisting of a containment condition annotation, a spatial proximity condition annotation, a color condition annotation, and a UMLS containment condition annotation, resulting in an annotated question;

selecting, by the processor, one or more candidate answers to the received question based on the lexical answer type;

generating, by the processor, one or more condition fulfillment scores of the one or more candidate answers based on comparing the one or more candidate answers against the one or more condition annotations included in the annotated question;

ranking, by the processor, the one or more candidate answers based upon the one or more condition fulfillment scores; and

providing the ranked one or more candidate answers to a user.

2. The method of claim 1 wherein, prior to discovering the one or more candidate answers, the annotating further comprises:

determining that the one or more terms correspond to the lexical answer type;

identifying the one or more syntactic relationships corresponding to the lexical answer type that indicate a presence of a containment condition corresponding to the lexical answer type; and

adding the one or more condition annotations to the question based upon the identified syntactic relationships.

3. The method of claim 1 further comprising:

selecting a first one of the one or more candidate answers and a first condition corresponding to a first one of the one or more condition annotations;

generating a first one of the one or more condition fulfillment scores by comparing the first candidate answer to the first condition;

selecting a second a second condition corresponding to a second one of the one or more condition annotations;

generating a second one of the one or more condition fulfillment scores by comparing the first candidate answer to the second condition;

aggregating the first condition fulfillment score and the second condition fulfillment score into an aggregated condition fulfillment score; and

assigning the aggregated condition fulfillment score to the first candidate answer.

4. The method of claim 1 further comprising:

traversing a knowledge base to identify one or more document passages corresponding to the lexical answer type; and

generating the one or more candidate answers based upon the identified one or more document passages.

5. The method of claim 1 wherein the lexical answer type corresponds to a word in the question that semantically represents an answer to the question.

6. An information handling system comprising:

one or more processors;

a memory coupled to at least one of the processors; and

a set of computer program instructions stored in the memory and executed by at least one of the processors in order to perform actions of;

receiving a question that includes a logical answer type;

identifying, by the processor, one or more terms in the question that impose one or more requirements of the question, wherein the one or more terms are different from the logical answer type;

annotating the question, by the processor, by adding one or more condition annotations to the question that correspond to the one or more requirements, wherein at least one of the one or more condition annotations are selected from the group consisting of a containment condition annotation, a spatial proximity condition annotation, a color condition annotation, and a UMLS containment condition annotation, resulting in an annotated question;

selecting, by the processor, one or more candidate answers to the received question based on the lexical answer type;

generating, by the processor, one or more condition fulfillment scores of the one or more candidate answers based on comparing the one or more candidate answers against the one or more condition annotations included in the annotated question;

ranking, by the processor, the one or more candidate answers based upon the one or more condition fulfillment scores; and

providing the ranked one or more candidate answers to a user.

7. The information handling system of claim 6 wherein, prior to discovering the one or more candidate answers, the one or more processors performs additional actions comprising:

determining that the one or more terms correspond to the lexical answer type;

identifying the one or more syntactic relationships corresponding to the lexical answer type that indicate a presence of a containment condition corresponding to the lexical answer type; and

adding the one or more condition annotations to the question based upon the identified syntactic relationships.

8. The information handling system of claim 6 wherein the one or more processors perform additional actions comprising:

selecting a first one of the one or more candidate answers and a first condition corresponding to a first one of the one or more condition annotations;

generating a first one of the one or more condition fulfillment scores by comparing the first candidate answer to the first condition;

selecting a second a second condition corresponding to a second one of the one or more condition annotations;

generating a second one of the one or more condition fulfillment scores by comparing the first candidate answer to the second condition;

aggregating the first condition fulfillment score and the second condition fulfillment score into an aggregated condition fulfillment score; and

assigning the aggregated condition fulfillment score to the first candidate answer.

9. The information handling system of claim 6 wherein the one or more processors perform additional actions comprising:

traversing a knowledge base to identify one or more document passages corresponding to the lexical answer type; and

generating the one or more candidate answers based upon the identified one or more document passages.

10. The information handling system of claim 6 wherein the lexical answer type corresponds to a word in the question that semantically represents an answer to the question.

11. A computer program product stored in a computer readable storage medium, comprising computer program code that, when executed by an information handling system, causes the information handling system to perform actions comprising:

receiving a question that includes a logical answer type;

identifying, by the processor, one or more terms in the question that impose one or more requirements of the question, wherein the one or more terms are different from the logical answer type;

annotating the question, by the processor, by adding one or more condition annotations to the question that correspond to the one or more requirements, wherein at least one of the one or more condition annotations are selected from the group consisting of a containment condition annotation, a spatial proximity condition annotation, a color condition annotation, and a UMLS containment condition annotation, resulting in an annotated question;

selecting, by the processor, one or more candidate answers to the received question based on the lexical answer type;

generating, by the processor, one or more condition fulfillment scores of the one or more candidate answers based on comparing the one or more candidate answers against the one or more condition annotations included in the annotated question;

ranking, by the processor, the one or more candidate answers based upon the one or more condition fulfillment scores; and

providing the ranked one or more candidate answers to a user.

12. The computer program product of claim 11 wherein, prior to discovering the one or more candidate answers, the one or more processors performs additional actions comprising:

determining that the one or more terms correspond to the lexical answer type;

identifying the one or more syntactic relationships corresponding to the lexical answer type that indicate a presence of a containment condition corresponding to the lexical answer type; and

adding the one or more condition annotations to the question based upon the identified syntactic relationships.

13. The computer program product of claim 11 wherein the one or more processors perform additional actions comprising:

selecting a first one of the one or more candidate answers and a first condition corresponding to a first one of the one or more condition annotations;

generating a first one of the one or more condition fulfillment scores by comparing the first candidate answer to the first condition;

selecting a second a second condition corresponding to a second one of the one or more condition annotations;

generating a second one of the one or more condition fulfillment scores by comparing the first candidate answer to the second condition;

aggregating the first condition fulfillment score and the second condition fulfillment score into an aggregated condition fulfillment score; and

assigning the aggregated condition fulfillment score to the first candidate answer.

14. The computer program product of claim 11 wherein the one or more processors perform additional actions comprising:

traversing a knowledge base to identify one or more document passages corresponding to the lexical answer type; and

generating the one or more candidate answers based upon the identified one or more document passages.