Identifying and displaying relationships between candidate answers

• US 9,558,264 B2
• Filed: 02/15/2016
• Issued: 01/31/2017
• Est. Priority Date: 12/05/2013
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Visotski, William G., Wilson, David E., Yates, Robert L.
• Primary Examiner(s): Jacob, Ajith
• Application Number:
  US15/043,785
• Publication Number:
  US 20160171095A1
• Time in Patent Office: 351 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/338 Presentation of query results
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri

Abstract

Mechanisms are provided for identifying commonalities between candidate answers generated by a Question and Answer (QA) system in response to an input question. The mechanisms receive a plurality of candidate answers for an input question from the QA system and identify terms present in the candidate answers. The mechanisms determine relationships between terms in each of the candidate answers and determine a common relationship between a first term and a second term, the common relationship being common amongst at least a subset of the plurality of candidate answers, based on the determined relationships between terms in each of the candidate answers. The mechanisms present the plurality of candidate answers and the common relationship to a user.

Claims

1. A method, in a data processing system, for identifying commonalities between candidate answers generated by a Question and Answer (QA) system in response to an input question, the method comprising:

receiving, by the data processing system, a plurality of candidate answers for an input question from the QA system;

identifying, by the data processing system, terms present in the candidate answers;

determining, by the data processing system, relationships between terms in each of the candidate answers;

determining, by the data processing system, a common relationship between a first term and a second term, the common relationship being common amongst at least a subset of the plurality of candidate answers, based on the determined relationships between terms in each of the candidate answers; and

presenting, by the data processing system, the plurality of candidate answers and the common relationship to a user.

2. The method of claim 1, wherein at least one of the first term or the second term is an entity having an entity type.

3. The method of claim 1, wherein determining the common relationship comprises determining an intersection of the relationships between terms in the subset.

4. The method of claim 1, wherein determining the common relationship comprises:

searching a relationship data store comprising a plurality of entries, each entry corresponding to a relationship between a first found term and at least one second found term found during pre-processing of documents of at least one corpus; and

identifying one or more entries in the relationship data store that match a term found in a candidate answer.

5. The method of claim 4, wherein each entry in the relationship data store comprises a pairwise relationship between a first term found during pre-processing of at least one document of a corpus, and a second term found during pre-processing of the at least one document of the corpus, based on relationships identified in the at least one document of the corpus through natural language processing of the at least one document of the corpus.

6. The method of claim 5, wherein each entry in the relationship data store further comprises a corresponding frequency of occurrence value indicating a frequency at which a corresponding pairwise relationship was found in the at least one document of the corpus during the pre-processing.

7. The method of claim 6, wherein presenting the plurality of candidate answers and the common relationship to a user comprises generating a visual display of the candidate answers and the relationships along with the corresponding frequency of occurrence value for each of the relationships.

8. The method of claim 7, wherein the visual display further comprises at least one evidence text passage from at least one document of a corpus that supports the common relationship, and wherein portions of at least one of the candidate answers or the evidence text passages are accentuated in the visual display based on a value corresponding to a frequency of occurrence value for the common relationship.

9. The method of claim 1, wherein determining a common relationship between the first term and the second term further comprises:

presenting a graphical user interface (GUI) providing a first portion of the GUI for outputting the candidate answers, a second portion of the GUI for outputting a plurality of user selectable options for specifying a desired common relationship that a user wishes to have identified between the candidate answers, a third portion of the GUI for outputting the relationships between terms in the candidate answers and the common relationship, and a fourth portion of the GUI for outputting evidence text passages from documents of a corpus that support the common relationship.

10. The method of claim 1, wherein presenting the plurality of candidate answers and the common relationship to a user comprises generating a visual display of the candidate answers and at least one evidence text passage from at least one document of a corpus that supports the common relationship, with portions of at least one of the candidate answers, or the at least one evidence text passage, being accentuated in the visual display.