Identification and verification of factual assertions in natural language

• US 9,483,582 B2
• Filed: 09/12/2014
• Issued: 11/01/2016
• Est. Priority Date: 09/12/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Gaucher, Brian P., Gil, Dario, Kephart, Jeffrey O., Lenchner, Jonathan, Melville, David O. S., Prager, John M., Salapura, Valentina
• Primary Examiner(s): Han, Qi
• Application Number:
  US14/484,489
• Publication Number:
  US 20160078149A1
• Time in Patent Office: 781 Days
• Field of Search: 704/235, 704/231, 704/251, 704/1, 704/4, 704/9, 704/257
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/19 Grammatical context, e.g. disambiguation of the recognition hypotheses based on word sequence rules
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence

Abstract

A system for verifying factual assertions in natural language. Receiving a text input. Identifying, a verifiable factual statement in the text. Forming a query based on the verifiable factual assertion. Searching a corpus based on the query. Determining the veracity of verifiable factual assertion based on the search results.

Claims

1. A method for verifying factual assertions in natural language, the method comprising:

monitoring a natural language input, wherein the natural language input includes an audio input;

processing the audio input into a digital input by an audio input processing module;

transmitting the digital input to a transcription module;

transcribing the digital input into a natural language text input by a speech transcription engine within the transcription module which matches the audio input to one or more acoustic models detailed by an electronic database;

identifying a phrase within the natural language text input, wherein the phrase comprises a subject, an object, and a relation;

determining if the phrase is a statement of opinion;

determining if the phrase is an indication of a consequence of a prior statement;

identifying the phrase as a factual assertion if the phase is not a statement of opinion and not an indication of a consequence of a prior statement;

verifying the factual assertion in the natural language text input; and

outputting a notification to a user of a result of the verification.

2. The method in accordance with claim 1, wherein:

determining if the phrase is a statement of opinion comprises identifying a predetermined word and/or word combinations indicative of a statement of opinion; and

determining if the phrase is an indication of a consequence of a prior statement comprises identifying a predetermined word and/or word combinations indicative of a consequence of a prior statement.

3. The method in accordance with claim 1, wherein verifying the factual assertion further comprises;

forming a question based on the factual assertion;

searching a corpus of information for an answer to the formed question; and

receiving a result of the search.

4. The method in accordance with claim 3, wherein forming the questions further comprises;

identifying relative pronouns in the factual assertion; and

substituting the relative pronouns with nouns and noun phrases.

5. The method in accordance with claim 1, wherein verifying the factual assertion further comprises;

comparing the result to the factual assertion, wherein the factual assertion is true when the result supports the factual assertion and the factual assertion is false if the result rejects the factual assertion and the factual assertion is inconclusive if the result neither supports or rejects the factual assertion.

6. The method in accordance with claim 5, wherein;

comparing a number of social media responses that support the factual assertion to a number of social media responses that reject the factual assertion, wherein the greater number of social media responses that support the factual assertion as compared to the number of social media responses that reject the factual assertion indicates a higher level of veracity.

7. The method in accordance with claim 1, further comprising:

determining if the factual assertion is ambiguous; and

resolving the ambiguity using contextual input.

8. A computer program product for verifying factual assertions in natural language, the computer program product comprising:

one or more computer readable storage media and program instructions stored on the one or more computer readable storage media, the program instructions comprising;

program instructions to monitor a natural language input, wherein the natural language input includes an audio input;

program instructions to process the audio input into a digital input by an audio input processing module;

program instructions to transmit the digital input to a transcription module;

program instructions to transcribe the digital input into a natural language text input by a speech transcription engine within the transcription module which matches the audio input to one or more acoustic models detailed by an electronic database;

program instructions to identify a phrase within the natural language text input, wherein the phrase comprises a subject, an object, and a relation;

program instructions to determine if the phrase is a statement of opinion;

program instructions to determine if the phrase is an indication of a consequence of a prior statement;

program instructions to identify the phrase as a factual assertion if the phase is not a statement of opinion and not an indication of a consequence of a prior statement;

program instructions to verify the factual assertion in the natural language text input; and

program instructions to output a notification to a user of a result of the verification.

9. The computer program product in accordance with claim 8, wherein:

program instructions to determine if the phrase is a statement of opinion comprises program instructions to identify a predetermined word and/or word combinations indicative of a statement of opinion; and

program instructions to determine if the phrase is an indication of a consequence of a prior statement comprises program instructions to identify a predetermined word and/or word combinations indicative of a consequence of a prior statement.

10. The computer program product in accordance with claim 8, wherein program instructions to verify the factual assertion further comprises;

program instructions to form a question based on the factual assertion;

program instructions to search a corpus of information for an answer to the formed question; and

program instructions to receive a result of the search.

11. The computer program product in accordance with claim 10, wherein program instructions to form the questions further comprises;

program instructions to identify relative pronouns in the factual assertion; and

program instructions to substitute the relative pronouns with nouns and noun phrases.

12. The computer program product in accordance with claim 8, wherein program instructions to verify the factual assertion further comprises;

program instructions to compare the result to the factual assertion, wherein the factual assertion is true when the result supports the factual assertion and the factual assertion is false if the result rejects the factual assertion and the factual assertion is inconclusive if the result neither supports or rejects the factual assertion.

13. The computer program product in accordance with claim 12, wherein;

program instructions to compare a number of social media responses that support the factual assertion to a number of social media responses that reject the factual assertion, wherein the greater number of social media responses that support the factual assertion as compared to the number of social media responses that reject the factual assertion indicates a higher level of veracity.

14. The computer program product in accordance with claim 8, further comprising:

program instructions to determine if the factual assertion is ambiguous; and

program instructions to resolve the ambiguity using contextual input.

15. A computer system for verifying factual assertions in natural language, the computer system comprising:

one or more computer processors;

one or more computer readable storage media;

program instructions stored on the computer readable storage media for execution by at least one of the one or more processors, the program instructions comprising;

program instructions to monitor a natural language input, wherein the natural language input includes an audio input;

program instructions to process the audio input into a digital input by an audio input processing module;

program instructions to transmit the digital input to a transcription module;

program instructions to transcribe the digital input into a natural language text input by a speech transcription engine within the transcription module which matches the audio input to one or more acoustic models detailed by an electronic database;

program instructions to identify a phrase within the natural language text input, wherein the phrase comprises a subject, an object, and a relation;

program instructions to determine if the phrase is a statement of opinion;

program instructions to determine if the phrase is an indication of a consequence of a prior statement;

program instructions to identify the phrase as a factual assertion if the phase is not a statement of opinion and not an indication of a consequence of a prior statement;

program instructions to verify the factual assertion in the natural language text input; and

program instructions to output a notification to a user of a result of the verification.

16. The computer system in accordance with claim 15, wherein:

program instructions to determine if the phrase is a statement of opinion comprises program instructions to identify a predetermined word and/or word combinations indicative of a statement of opinion; and

program instructions to determine if the phrase is an indication of a consequence of a prior statement comprises program instructions to identify a predetermined word and/or word combinations indicative of a consequence of a prior statement.

17. The computer system in accordance with claim 15, wherein program instructions to verify the factual assertion further comprises;

program instructions to form a question based on the factual assertion;

program instructions to search a corpus of information for an answer to the formed question; and

program instructions to receive a result of the search.

18. The computer system in accordance with claim 17, wherein program instructions to form the questions further comprises;

program instructions to identify relative pronouns in the factual assertion; and

program instructions to substitute the relative pronouns with nouns and noun phrases.

19. The computer system in accordance with claim 15, wherein program instructions to verify the factual assertion further comprises;

program instructions to compare the result to the factual assertion, wherein the factual assertion is true when the result supports the factual assertion and the factual assertion is false if the result rejects the factual assertion and the factual assertion is inconclusive if the result neither supports or rejects the factual assertion.