Social media data mining for early detection of newsworthy civil unrest events

• US 9,953,080 B1
• Filed: 06/03/2014
• Issued: 04/24/2018
• Est. Priority Date: 04/05/2013
• Status: Active Grant

• Current Assignee: HRL Laboratories LLC (The Boeing Co.)
• Original Assignee: HRL Laboratories LLC (The Boeing Co.)
• Inventors: Compton, Ryan F.
• Primary Examiner(s): HASAN, SYED HAROON
• Application Number:
  US14/295,101
• Time in Patent Office: 1,421 Days
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities
• CPC Class Codes: G06Q 10/101 Collaborative creation, e.g. joint development of products or services
• CPC Class Codes: G06Q 50/26 Government or public services business processes related to the transportation industry G06Q50/40

Abstract

Described is system for early detection of civil unrest events. A set of textual social media data having corresponding user identifications is received as input. The textual social media data is filtered with a keyword filter, resulting in a first set of data comprising keywords of interest. The first set of data is then filtered with a future date filter, resulting in a second set of data comprising future dates. The second set of data is filtered based on relevance to a civil unrest event with a text classifier module, resulting in a third set of data that is classified based on relevance to the civil unrest event. The third set of data is then filtered based on a location of interest, resulting in a fourth set of data having user identifications in the location of interest. A location is assigned to the civil unrest event, and a forecast of the civil unrest event is output.

Claims

1. A system for early detection of civil unrest events, the system comprising:

one or more processors and a non-transitory computer-readable medium having executable instructions encoded thereon such that when executed, the one or more processors perform operations of;

receiving a set of textual social media data having corresponding user identifications as input;

filtering the set of textual social media data with a keyword filter, resulting in a first set of data comprising keywords of interest;

filtering the first set of data with a future date filter, resulting in a second set of data comprising future dates;

filtering the second set of data based on relevance to a civil unrest event with a text classifier module, resulting in a third set of data that is classified based on relevance to the civil unrest event;

filtering the third set of data based on a location of interest, resulting in a fourth set of data having user identifications in the location of interest;

assigning a location to the civil unrest event; and

outputting a forecast of the civil unrest event.

2. The system as set forth in claim 1, wherein the forecast comprises a demographic of participants of the civil unrest event, a reason for the civil unrest event, a date the civil unrest event is forecast to occur on, a location where the civil unrest event is expected to occur, and a probability of how likely it is that the civil unrest event will occur.

3. The system as set forth in claim 2, wherein the text classifier module comprises a standard logistic regression classifier trained on the set of textual social media data.

4. The system as set forth in claim 3, wherein the text classifier module further comprises a user type identification classifier to determine whether a user identification is related to an individual or an organization, and wherein the one or more processors further perform an operation of increasing the probability of the civil unrest event occurring if the user identification is related to an organization.

5. The system as set forth in claim 4, wherein the one or more processors further perform an operation of searching the set of textual social media data for mentions of cities or monuments to assign a location to the civil unrest event.

6. The system as set forth in claim 5, wherein if there are no mentions of cities or monuments in the set of textual social media data, the one or more processors further perform an operation of assigning a location to the civil unrest event using a geocoding module that infers a location of a user identification based on a set of social media relationships within the set of textual social media data.

19. The system as set forth in claim 1, wherein the one or more processors further perform an operation of assigning a probability score to the civil unrest event.

20. The system as set forth in claim 1, wherein the one or more processors further perform an operation of condensing duplicate forecasts into one forecast by averaging probabilities of how likely it is that the civil unrest event will occur.

21. The system as set forth in claim 1, wherein the text classifier module further comprises a user type identification classifier to determine whether a user identification is related to an individual or an organization, and wherein the one or more processors further perform an operation of increasing the probability of the civil unrest event occurring if the user identification is related to an organization.

22. The system as set forth in claim 1, wherein the one or more processors further perform an operation of searching the set of textual social media data for mentions of cities or monuments to assign a location to the civil unrest event.

23. The system as set forth in claim 1, wherein if there are no mentions of cities or monuments in the set of textual social media data, the one or more processors further perform an operation of assigning a location to the civil unrest event using a geocoding module that infers a location of a user identification based on a set of social media relationships within the set of textual social media data.

7. A computer-implemented method for early detection of civil unrest events, comprising:

an act of causing a data processor to execute instructions stored on a non-transitory memory such that upon execution, the data processor performs operations of;

receiving a set of textual social media data having corresponding user identifications as input;

filtering the set of textual social media data with a keyword filter, resulting in a first set of data comprising keywords of interest;

filtering the first set of data with a future date filter, resulting in a second set of data comprising future dates;

filtering the second set of data based on relevance to a civil unrest event with a text classifier module, resulting in a third set of data that is classified based on relevance to the civil unrest event;

filtering the third set of data based on a location of interest, resulting in a fourth set of data having user identifications in the location of interest;

assigning a location to the civil unrest event; and

outputting a forecast of the civil unrest event.

8. The method as set forth in claim 7, wherein the forecast comprises a demographic of participants of the civil unrest event, a reason for the civil unrest event, a date the civil unrest event is forecast to occur on, a location where the civil unrest event is expected to occur, and a probability of how likely it is that the civil unrest event will occur.

9. The method as set forth in claim 8, wherein the text classifier module comprises a standard logistic regression classifier trained on the set of textual social media data.

10. The method as set forth in claim 9, wherein the text classifier module further comprises a user type identification classifier to determine whether a user identification is related to an individual or an organization, and wherein the one or more processors further perform an operation of increasing the probability of the civil unrest event occurring if the user identification is related to an organization.

11. The method as set forth in claim 10, wherein the one or more processors further perform an operation of searching the set of textual social media data for mentions of cities or monuments to assign a location to the civil unrest event.

12. The method as set forth in claim 11, wherein if there are no mentions of cities or monuments in the set of textual social media data, the one or more processors further perform an operation of assigning a location to the civil unrest event using a geocoding module that infers a location of a user identification based on a set of social media relationships within the set of textual social media data.

13. A computer program product for early detection of civil unrest events, the computer program product comprising computer-readable instructions stored on a non-transitory computer-readable medium that are executable by a computer having a processor for causing the processor to perform operations of:

receiving a set of textual social media data having corresponding user identifications as input;

filtering the set of textual social media data with a keyword filter, resulting in a first set of data comprising keywords of interest;

filtering the first set of data with a future date filter, resulting in a second set of data comprising future dates;

filtering the second set of data based on relevance to a civil unrest event with a text classifier module, resulting in a third set of data that is classified based on relevance to the civil unrest event;

filtering the third set of data based on a location of interest, resulting in a fourth set of data having user identifications in the location of interest;

assigning a location to the civil unrest event; and

outputting a forecast of the civil unrest event.

14. The computer program product as set forth in claim 13, wherein the forecast comprises a demographic of participants of the civil unrest event, a reason for the civil unrest event, a date the civil unrest event is forecast to occur on, a location where the civil unrest event is expected to occur, and a probability of how likely it is that the civil unrest event will occur.

15. The computer program product as set forth in claim 14, wherein the text classifier module comprises a standard logistic regression classifier trained on the set of textual social media data.

16. The computer program product as set forth in claim 15, wherein the text classifier module further comprises a user type identification classifier to determine whether a user identification is related to an individual or an organization, and wherein the one or more processors further perform an operation of increasing the probability of the civil unrest event occurring if the user identification is related to an organization.

17. The computer program product as set forth in claim 16, further comprising instructions for causing the processor to perform an operation of searching the set of textual social media data for mentions of cities or monuments to assign a location to the civil unrest event.

18. The computer program product as set forth in claim 17, wherein if there are no mentions of cities or monuments in the set of textual social media data, the computer program product further comprises instructions for causing the processor to perform an operation of assigning a location to the civil unrest event using a geocoding module that infers a location of a user identification based on a set of social media relationships within the set of textual social media data.