Tracking and prediction of societal event trends using amplified signals extracted from social media

• US 9,892,168 B1
• Filed: 05/24/2016
• Issued: 02/13/2018
• 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: Apreleva, Sofia S. V., Lu, Tsai-Ching, Compton, Ryan F., Keegan, Matthew S., Lee, Craig
• Primary Examiner(s): Hasan, Syed
• Application Number:
  US15/163,547
• Time in Patent Office: 630 Days
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/2477 Temporal data queries
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

Described is a system for tracking and predicting social events. The system filters a time series of data obtained from a social media source. Enhanced filtered signals (EFS) are extracted from the filtered time series data based on an amplification signal obtained via a summation of signals relevant to a process of interest in the filtered time series data. A level of human social activity in the social media source is monitored by comparing the extracted EFS to an event database to detect an increase in a number of social activity events in the social media source compared to the event database.

Claims

1. A system for monitoring social events using signals extracted from a social media platform, the system comprising:

one or more processors and a non-transitory memory having instructions encoded thereon such that when the instructions are executed, the one or more processors perform operations of;

filtering a time series of social network data obtained from a social media platform based on a correlation between the time series and a set of around truth data to remove noise from the time series;

extracting enhanced filtered signals (EFS) from the filtered time series of social network data based on an amplification of a signal obtained via a summation of signals relevant to a social activity event in the filtered time series of social network data;

monitoring a level of human social activity related to the social activity event in the social network data;

detecting an increase in a number of social activity events in the social network data compared to the set of around truth data; and

generating an alert regarding the increase in the number of social activity events, wherein the alert is provided electronically via an electronic system.

2. The system as set forth in claim 1, wherein the extracted EFS is used to generate at least one prediction of at least one of a future social activity event and a trend in human social activity related to the future social activity event.

3. The system as set forth in claim 1, wherein the one or more processors further perform an operation of filtering the time series of data with a threshold value T_r, wherein the threshold value T_r is an average correlation coefficient between a randomly generated time series of data and the set of ground truth data.

4. The system as set forth in claim 3, wherein the one or more processors further perform an operation of regressing the extracted EFS on the set of ground truth data with different time windows to obtain a set of regression fits between the extracted EFS and the set of ground truth data.

5. The system as set forth in claim 4, wherein the one or more processors further perform operations of:

obtaining a set of parameters from the set of regression fits; and

generating the at least one prediction of the future social activity event and the trend in human social activity related to the social activity event by applying the set of parameters to new sets of time series data.

6. The system as set forth in claim 2, wherein Bayesian model averaging is used to weight at least one of the prediction of the future social activity event and the trend in human social activity related to the future social activity event according to historical data.

7. A computer-implemented method for monitoring social events, comprising:

an act of causing one or more processors to execute instructions stored on a non-transitory memory such that upon execution, the one or more processors perform operations of;

filtering a time series of social network data obtained from a social media platform based on a correlation between the time series and a set of ground truth data to remove noise from the time series;

extracting enhanced filtered signals (EFS) from the filtered time series of social network data based on an amplification of a signal obtained via a summation of signals relevant to a social activity event in the filtered time series of social network data;

monitoring a level of human social activity related to the social activity event in the social network data;

detecting an increase in a number of social activity events in the social network data compared to the set of ground truth data; and

generating an alert regarding the increase in the number of social activity events, wherein the alert is provided electronically via an electronic system.

8. The method as set forth in claim 7, wherein the extracted EFS is used to generate at least one prediction of at least one of a future social activity event and a trend in human social activity related to the future social activity event.

9. The method as set forth in claim 7, wherein the one or more processors further perform an operation of filtering the time series of data with a threshold value T_r, wherein the threshold value T_r is an average correlation coefficient between a randomly generated time series of data and the set of ground truth data.

10. The method as set forth in claim 9, wherein the one or more processors further perform an operation of regressing the extracted EFS on the set of ground truth data with different time windows to obtain a set of regression fits between the extracted EFS and the set of ground truth data.

11. The method as set forth in claim 10, wherein the one or more processors further perform operations of:

obtaining a set of parameters from the set of regression fits; and

generating at least one of the prediction of the future social activity event and the trend in human social activity related to the future social activity event by applying the set of parameters to new sets of time series data.

12. The method as set forth in claim 8, wherein Bayesian model averaging is used to weight at least one of the prediction of the future social activity event and the trend in human social activity related to the future social activity event according to historical data.

13. A computer program product for monitoring social 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:

filtering a time series of social network data obtained from a social media platform based on a correlation between the time series and a set of ground truth data to remove noise from the time series;

extracting enhanced filtered signals (EFS) from the filtered time series of social network data based on an amplification of a signal obtained via a summation of signals relevant to a social activity event in the filtered time series of social network data;

monitoring a level of human social activity related to the social activity event in the social network data;

detecting an increase in a number of social activity events in the social network data compared to the set of around truth data; and

generating an alert regarding the increase in the number of social activity events, wherein the alert is provided electronically via an electronic system.

14. The computer program product as set forth in claim 13, wherein the extracted EFS is used to generate at least one prediction of at least one of a future social activity event and a trend in human social activity related to the future social activity event.

15. The computer program product as set forth in claim 13, further comprising instructions for causing the one or more processors to perform an operation of filtering the time series of data with a threshold value T_r, wherein the threshold value T_r is an average correlation coefficient between a randomly generated time series of data and the set of ground truth data.

16. The computer program product as set forth in claim 15, further comprising instructions for causing the one or more processors to perform an operation of regressing the extracted EFS on the set of ground truth data with different time windows to obtain a set of regression fits between the extracted EFS and the set of ground truth data.

17. The computer program product as set forth in claim 16, further comprising instructions for causing the one or more processors to perform operations of:

obtaining a set of parameters from the set of regression fits; and

generating at least one of the prediction of the future social activity event and the trend in human social activity related to the social activity event by applying the set of parameters to new sets of time series data.

18. The computer program product as set forth in claim 14, wherein Bayesian model averaging is used to weight at least one of the prediction of the future social activity event and the trend in human social activity related to the social activity event according to historical data.