Corpus augmentation system

• US 9,940,370 B2
• Filed: 01/02/2015
• Issued: 04/10/2018
• Est. Priority Date: 01/02/2015
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bastide, Paul R., Broomhall, Matthew E., Cohen, Aaron M., Desforges, Christopher W., Kraenzel, Carl J., Loredo, Robert E.
• Primary Examiner(s): Khong, Alexander
• Application Number:
  US14/588,719
• Publication Number:
  US 20160196312A1
• Time in Patent Office: 1,194 Days
• Field of Search: None
• CPC Class Codes: G06F 16/2379 Updates performed during online database operations; commit processing
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

An approach is provided for automatically ingesting additional corpus based on an interaction history that is mined to identify a question that meets specified answer deficiency criteria, and then generate a second question which is correlated to the first question by requesting additional answer information for answering the first question, where the second question is posted to a forum using a selected persona so that forum responses can be monitored and ingested as additional content in the knowledge base.

Claims

1. An information handling system comprising:

one or more processors;

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

a set of instructions stored in the memory and executed by at least one of the processors to ingest additional content in a knowledge base, wherein the set of instructions are executable to perform actions of;

mining, by the system, an interaction history comprising a plurality of questions and answer results to identify a first question by performing a natural language processing (NLP) analysis of the plurality of questions and answer results to detect the first question that meets specified answer deficiency criteria;

generating, by the system, a second question which is correlated to the first question by extracting a text sentence from one or more documents correlated to the first question and parsing the text sentence to populate a defined question template used to construct the second question requesting additional answer information for answering the first question;

selecting, by the system, at least one persona to post the second question;

posting, by the system, the second question to a forum using the at least one persona;

monitoring, by the system, the forum for responses to the second question; and

ingesting, by the system, any response to the second question as additional content in the knowledge base.

2. The information handling system of claim 1, wherein the NLP analysis identifies the first question by detecting an answer for the first question that has a confidence measure below a minimum confidence threshold, by detecting an answer for the first question that provides no response, by detecting an answer for the first question that has an associated negative sentiment, or by detecting an answer for the first question that has no supporting evidence.

3. The information handling system of claim 1, wherein generating the second question comprises:

retrieving, by the system, one or more documents associated with the identified first question;

extracting, by the system, a text sentence from the one or more documents which is correlated to the first question; and

generating, by the system, the second question by parsing the text sentence to populate a defined question template used to construct the second question.

4. The information handling system claim 1, wherein selecting the at least one persona comprises selecting a persona from a group consisting of a primed persona, a curated persona, an automated persona, and a selected persona.

5. The information handling system claim 1, wherein the at least one persona is registered at the forum where the second question is posted.

6. The information handling system claim 1, further comprising an additional set of instructions stored in the memory and executed by at least one of the processors to perform action of generating, by the system, one or more persona associated with a selected forum by extracting user profile information from users registered with the selected forum.

7. The information handling system claim 1, further comprising an additional set of instructions stored in the memory and executed by at least one of the processors to perform action of generating, by the system, one or more question templates associated with a selected forum by using machine learning processing to identify one or more questions from users registered with the selected forum and extract therefrom one or more question templates.

8. A computer program product stored in a computer readable storage medium, comprising computer instructions that, when executed by an information handling system, causes the system to ingest additional content in a knowledge base by performing actions comprising:

mining, by the system, an interaction history comprising a plurality of questions and answer results to identify a first question by performing a natural language processing (NLP) analysis of the plurality of questions and answer results to detect the first question that meets specified answer deficiency criteria;

generating, by the system, a second question which is correlated to the first question by extracting a text sentence from one or more documents correlated to the first question and parsing the text sentence to populate a defined question template used to construct the second question requesting additional answer information for answering the first question;

selecting, by the system, at least one persona to post the second question;

posting, by the system, the second question to a forum using the at least one persona;

monitoring, by the system, the forum for responses to the second question; and

ingesting, by the system, any response to the second question as additional content in the knowledge base.

9. The computer program product of claim 8, wherein mining the interaction history comprises performing, by the system, identifying the first question by detecting an answer for the first question that has a confidence measure below a minimum confidence threshold, by detecting an answer for the first question that provides no response, by detecting an answer for the first question that has an associated negative sentiment, or by detecting an answer for the first question that has no supporting evidence.

10. The computer program product of claim 8, wherein the NLP analysis identifies the first question by detecting an answer for the first question that has a confidence measure below a minimum confidence threshold, by detecting an answer for the first question that provides no response, by detecting an answer for the first question that has an associated negative sentiment, or by detecting an answer for the first question that has no supporting evidence.

11. The computer program product of claim 8, wherein selecting the at least one persona comprises selecting a persona from a group consisting of a primed persona, a curated persona, an automated persona, and a selected persona.

12. The computer program product of claim 8, further comprising computer instructions that are executed to cause the system to generate one or more persona associated with a selected forum by extracting user profile information from users registered with the selected forum.

13. The computer program product of claim 8, further comprising computer instructions that are executed to cause the system to generate one or more question templates associated with a selected forum by using machine learning processing to identify one or more questions from users registered with the selected forum and extract therefrom one or more question templates.