System for identifying, monitoring and ranking incidents from social media

• US 9,397,904 B2
• Filed: 12/30/2013
• Issued: 07/19/2016
• Est. Priority Date: 12/30/2013
• Status: Active Grant

• Current Assignee: AirBNB Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Berlingerio, Michele, Dong, Xiaowen, Gkoulalas-Divanis, Aris, Mavroeidis, Dimitrios
• Primary Examiner(s): Lewis, Cheryl
• Application Number:
  US14/143,949
• Publication Number:
  US 20150186378A1
• Time in Patent Office: 932 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04L 12/1813 for computer conferences, e.g. chat rooms instant messaging H04L51/04; protocols for multimedia communication H04L65/1101; arrangements for multi-party communication H04L65/403; telephonic conference arrangements H04M3/56; television conference systems H04N7/15
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation
• CPC Class Codes: H04L 51/23 Reliability checks, e.g. acknowledgments or fault reporting
• CPC Class Codes: H04L 51/52 for supporting social networking services

Abstract

A system and method for detecting, monitoring and ranking incidents from social media streams comprises detecting incidents from social media streams and continuously monitoring the incidents, calculating a current-score for each of the incidents, determining a projected-score indicating an expected evolution for each of the incidents, ranking the incidents based on the current-scores of the incidents, predicting ranking of the incidents based on the projected-score of the incidents, and updating the predicted ranking responsive to new input detected about the incidents from the social media streams. In one aspect, the current-score is computed in accordance with characteristics of the incident comprising one or more of social impact assessment, dynamic location of users, human perception, and social network features. In one aspect, the current-score is calculated using one or more diffusion models based on one of a type of the incident, and similar incidents for which their evolution is known.

Claims

1. A method for detecting, monitoring and ranking incidents from social media streams, the method comprising:

detecting one or more incidents from social media streams and continuously monitoring the one or more incidents;

calculating a current-score for each of the one or more incidents;

determining a projected-score indicating an expected evolution for each of the one or more incidents, and using a finished incidents for determining the expected evolution for said each of the one or more incidents;

ranking the incidents based on the current-scores of the incidents;

predicting, using a critically based incident ranker, a ranking of the incidents based on the projected-score of the incidents; and

updating the predicted ranking responsive to a new input detected about the incidents from the social media streams,

wherein a hardware processor device is configured to perform said detecting one or more

incidents, calculating a current-score, determining a projected-score, ranking the incidents,

predicting ranking of the incidents and updating the predicted ranking.

2. The method as in claim 1, wherein the current-score is computed in accordance with characteristics of the incident comprising one or more of social impact assessment, dynamic location of users, human perception, and social network features.

3. The method as in claim 1, wherein the current-score is calculated using one or more diffusion models based on one of a type of the incident, and similar incidents having known evolution.

4. The method as in claim 3, wherein updating the predicted ranking comprises at least a user providing real-time feedback on the incident, said feedback influencing prediction of current state of the incident, and said feedback used with the one or more diffusion models used to calculate the current-score, to update the expected evolution of the incident.

5. The method as in claim 3, wherein updating the predicted ranking is performed automatically based on the expected evolution of the incident based on the one or more diffusion models.

6. The method as in claim 1, further comprising displaying the predicted ranking on a display device.

7. A system for detecting, monitoring and ranking incidents from social media streams, comprising:

a memory device;

a display device; and

a hardware processor coupled to the memory device, the processor configured to;

detect one or more incidents from social media streams and continuously monitor the one or more incidents;

calculate a current-score for each of the one or more incidents;

determine a projected-score indicating an expected evolution for each of the one or more

incidents, and using a finished incidents for determining the expected evolution for said each of the one or more incidents;

rank the incidents based on the current-scores of the incidents;

predict, using a critically based incident ranker, a ranking of the incidents based on the projected-score of the incidents; and

update the predicted ranking responsive to a new input detected about the incidents from the social media streams.

8. The system as in claim 7, wherein the current-score is computed in accordance with characteristics of the incident comprising one or more of social impact assessment, dynamic location of users, human perception, and social network features.

9. The system as in claim 7, wherein the current-score is calculated using one or more diffusion models based on one of a type of the incident, and similar incidents having known evolution.

10. The system as in claim 9, wherein updating the predicted ranking comprises at least a user providing real-time feedback on the incident, said feedback influencing prediction of current state of the incident, and said feedback used with the one or more diffusion models used to calculate the current-score, to update the expected evolution of the incident.

11. The system as in claim 9, wherein update the predicted ranking is performed automatically based on the expected evolution of the incident based on the one or more diffusion models.

12. The system as in claim 7, further comprising a display device configured to display the predicted ranking.

13. A computer program product for detecting, monitoring and ranking incidents from social media streams, the computer program product comprising a non-transitory storage medium readable by a processing circuit and storing instructions run by the processing circuit for performing method steps for dynamic, semi-supervised clustering executed on a hardware processor, comprises:

detecting one or more incidents from social media streams and continuously monitoring the one or more incidents;

calculating a current-score for each of the one or more incidents;

determining a projected-score indicating an expected evolution for each of the one or more incidents, and using a finished incidents for determining the expected evolution for said each of the one or more incidents;

ranking the incidents based on the current-scores of the incidents;

predicting, using a critically based incident ranker, a ranking of the incidents based on the projected-score of the incidents; and

updating the predicted ranking responsive to a new input detected about the incidents from the social media streams.

14. The computer program product as in claim 13, wherein the current-score is computed in accordance with characteristics of the incident comprising one or more of social impact assessment, dynamic location of users, human perception, and social network features.

15. The computer program product as in claim 13, wherein the current-score is calculated using one or more diffusion models based on one of a type of the incident, and similar incidents having known evolution.

16. The computer program product as in claim 15, wherein updating the predicted ranking comprises at least a user providing real-time feedback on the incident, said feedback influencing prediction of current state of the incident, and said feedback used with the one or more diffusion models used to calculate the current-score, to update the expected evolution of the incident.

17. The computer program product as in claim 15, wherein updating the predicted ranking is performed automatically based on the expected evolution of the incident based on the one or more diffusion models.

18. The computer program product as in claim 13, further comprising displaying the predicted ranking on a display device.