Optimized narrative generation and fact checking method and system based on language usage

• US 9,454,562 B2
• Filed: 05/01/2015
• Issued: 09/27/2016
• Est. Priority Date: 09/04/2014
• Status: Active Grant

• Current Assignee: Lucas Myslinski
• Original Assignee: Lucas Myslinski
• Inventors: Myslinski, Lucas J.
• Primary Examiner(s): HUANG, MIRANDA M
• Application Number:
  US14/701,727
• Publication Number:
  US 20160070742A1
• Time in Patent Office: 515 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2365 Ensuring data consistency and integrity
• CPC Class Codes: G06F 16/244 Grouping and aggregation
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06F 16/335 Filtering based on additional data, e.g. user or group profiles filtering in web context G06F16/9535, G06F16/9536
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 40/226 Validation
• CPC Class Codes: G06F 40/237 Lexical tools
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 50/01 Social networking

Abstract

An optimized fact checking system analyzes and determines the factual accuracy of information and/or characterizes the information by comparing the information with source information. The optimized fact checking system automatically monitors information, processes the information, fact checks the information in an optimized manner and/or provides a status of the information. In some embodiments, the optimized fact checking system generates, aggregates, and/or summarizes content.

Claims

1. A method programmed in a non-transitory memory of a device comprising:

a. automatically generating narrative information by analyzing at least one first data source including extracting a first set of data from the at least one first data source and organizing the first set of data from the at least one first data source into a narrative, wherein the first set of data includes a plurality of different perspectives, wherein the plurality of different perspectives are based on a plurality of different languages, further wherein a first perspective of the plurality of perspectives is a first content in a first language, and a second perspective of the plurality of perspectives is a second content in a second language, wherein the first language and the second language are different;

b. automatically fact checking, with the device, the narrative information by comparing the narrative information with source information to generate a result, wherein the narrative information includes the first content in the first language and the second content in the second language, wherein fact checking the first content comprises fact checking using sources in the second language translated into the first language, and fact checking the second content translated into the first language comprises fact checking using sources in the first language; and

c. automatically providing a status of the narrative information in real-time based on the result of the comparison of the narrative information with the source information.

2. The method of claim 1 wherein generating the narrative information utilizes social networking information of a user to determine a focus of the at least one first data source based on a keyword detected most frequently, wherein the user is being provided the status of the narrative information.

3. The method of claim 1 wherein generating the narrative information utilizes one or more data sources with opposing perspectives, wherein the at least one first data source comprises a data source with a first opposing perspective, and the at least one first data source comprises a second data source with a second opposing perspective, wherein generating the narrative information includes generating a plurality of summaries with opposing perspectives and combining the summaries to form the narrative information such that a first sentence from a first summary of the plurality of summaries is provided first, a first sentence from a second summary of the plurality of summaries is provided second, and continuing in order until a last sentence from a last summary of the plurality of summaries is provided last.

4. The method of claim 1 further comprising automatically generating a template based on entity detection including retrieving text and placeholders from a data structure based on the entity detection, wherein generating the narrative information utilizes the template.

5. The method of claim 1 wherein the at least one first data source comprises aggregate information which comprises a plurality of articles about a same topic determined by comparing dates and matching keywords within the plurality of articles.

6. The method of claim 1 further comprising summarizing the narrative information using lexical chaining and the result generated by fact checking the narrative information, including determining lexical chain strengths, wherein a total strength of a sentence of the narrative information is based on a lexical chain strength and a factual accuracy of the sentence, further wherein the total strength of the sentence is increased when the sentence is factually accurate, and the total strength of the sentence is decreased when the sentence is factually inaccurate.

7. The method of claim 1 further comprising summarizing the source information using lexical chaining to generate summarized source information, and using the summarized source information for a preliminary comparison of the narrative information with the source information, and then using full source information corresponding to the summarized source information for the comparison to generate the result.

8. The method of claim 1 wherein generating the narrative information includes analyzing a validity rating of an entity, wherein the validity rating is based on factual accuracy of content provided by the entity, and only extracting data as the first set of data from the content provided by the entity when the validity rating of the entity is above a validity rating threshold.

9. The method of claim 1 wherein generating the narrative information includes analyzing a satisfaction rating of an entity, wherein the satisfaction rating is based on satisfaction analysis of content provided by the entity, and only extracting data as the first set of data from the content provided by the entity when the satisfaction rating of the entity is above a satisfaction rating threshold.

10. The method of claim 1 further comprising summarizing the narrative information, wherein summarizing the narrative information includes determining a set of most factually accurate sentences combined with lexical chaining to generate the summary.

11. The method of claim 1 wherein generating the narrative information by analyzing the at least one first data source includes matching user data with a template classification to select a template specific to the user, wherein the template specific to the user includes the at least one first data source specific to the user, wherein the template specific to the user is utilized in organizing the first set of data from the at least one first data source into the narrative.

12. The method of claim 1 further comprising generating advertisement information based on a summary of the narrative information, further wherein keywords of a lexical chain are analyzed to generate the advertisement information.

13. The method of claim 1 wherein the at least one first data source includes augmented reality-acquired device information acquired using an augmented reality device, and wherein the status of the narrative information is displayed on the augmented reality device.

14. The method of claim 1 wherein generating the narrative information includes replacing one or more placeholders within a template with extracted data.

15. A method programmed in a non-transitory memory of a device comprising:

a. automatically aggregating information utilizing a web crawler to search for and retrieve content to generate aggregated information;

b. automatically generating narrative information by analyzing at least one first data source including extracting a first set of data from the at least one first data source and organizing the first set of data from the at least one first data source into a narrative, wherein the at least one first data source comprises the aggregated information;

c. automatically fact checking, with the device, the narrative information by comparing the narrative information with source information to generate a result, wherein the narrative information includes a first content in a first language and a second content in a second language, wherein fact checking the first content comprises fact checking using sources in the second language translated into the first language, and fact checking the second content translated into the first language comprises fact checking using sources in the first language; and

d. automatically providing a status of the narrative information in real-time based on the result of the comparison of the narrative information with the source information.

16. A device comprising:

a. a non-transitory memory for storing an application for automatically performing the following steps;

i. generating narrative information by analyzing at least one first data source including extracting a first set of data from the at least one first data source and organizing the first set of data from the at least one first data source into a narrative, wherein generating the narrative information by analyzing the at least one first data source includes matching user data with a template classification to select a template specific to the user, wherein the template specific to the user includes the at least one first data source specific to the user, wherein the template specific to the user is utilized in organizing the first set of data from the at least one first data source into the narrative, wherein the first set of data includes a plurality of different perspectives, wherein the plurality of different perspectives are based on a plurality of different languages, further wherein a first perspective of the plurality of perspectives is a first content in a first language, and a second perspective of the plurality of perspectives is a second content in a second language, wherein the first language and the second language are different;

ii. fact checking the narrative information by comparing the narrative information with source information to generate a fact check result, wherein the narrative information includes the first content in the first language and the second content in the second language, wherein fact checking the first content comprises fact checking using sources in the second language translated into the first language, and fact checking the second content translated into the first language comprises fact checking using sources in the first language; and

iii. providing the narrative information and the fact check result in real-time; and

b. a processor for processing the application.

17. The device of claim 16 further comprising generating aggregate information, and the at least one first data source comprises the aggregate information.

18. The device of claim 17 further comprising modifying the aggregate information to generate modified aggregate information by deleting sentences containing factually inaccurate content in the aggregate information as determined by fact checking, wherein generating the narrative information utilizes the modified aggregate information.

19. The device of claim 17 wherein the aggregate information includes a plurality of articles and each article of the plurality of articles includes a plurality of sentences, wherein when a sentence is determined to be factually accurate based on fact checking and relevant to a summary based on lexical chaining, then the sentence is utilized when generating the summary, and when the sentence is determined to be factually inaccurate based on fact checking and relevant to the summary based on lexical chaining, then the sentence is not utilized when generating the summary, and when the sentence is determined to be factually accurate based on fact checking and not relevant to the summary based on lexical chaining, then the sentence is not utilized when generating the summary, and when the sentence is determined to be factually inaccurate based on fact checking and not relevant to the summary based on lexical chaining, then the sentence is not utilized when generating the summary.

20. The device of claim 17 wherein the aggregate information includes a plurality of articles, and each article of the plurality of articles receives a factual accuracy rating based on fact checking, further wherein an article of the plurality of articles with a lowest factual accuracy rating is summarized with a base number of sentences, and articles with higher factual accuracy ratings are summarized with the base number of sentences plus additional sentences such that the article with the highest factual accuracy rating is summarized with the base number of sentences plus a maximum number of additional sentences.