Dynamic event detection system and method

• US 10,678,815 B2
• Filed: 04/13/2017
• Issued: 06/09/2020
• Est. Priority Date: 10/02/2012
• Status: Active Grant

• Current Assignee: Banjo, Inc.
• Original Assignee: Banjo, Inc.
• Inventors: Patton, Damien, Mehta, Rish, Alves, Pedro
• Primary Examiner(s): Alam, Shahid A
• Application Number:
  US15/486,978
• Publication Number:
  US 20170220668A1
• Time in Patent Office: 1,153 Days
• Field of Search: 707740
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/487 using geographical or spatial information, e.g. location
• CPC Class Codes: G06F 16/489 using time information
• CPC Class Codes: G06F 16/5866 using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
• CPC Class Codes: H03F 1/0288 using a main and one or several auxiliary peaking amplifiers whereby the load is connected to the main amplifier using an impedance inverter, e.g. Doherty amplifiers
• CPC Class Codes: H03F 1/3247 using feedback acting on predistortion circuits H03F1/3264 takes precedence
• CPC Class Codes: H03F 2201/3227 Adaptive predistortion based on amplitude, envelope or power level feedback from the output of the main amplifier
• CPC Class Codes: H04B 1/0475 with means for limiting noise, interference or distortion H04B1/0483 takes precedence
• CPC Class Codes: H04B 2001/045 with means for improving efficiency
• CPC Class Codes: H04W 4/021 Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
• CPC Class Codes: H04W 4/21 for social networking applications
• CPC Class Codes: H04W 52/346 distributing total power among users or channels
• CPC Class Codes: H04W 52/52 using AGC [Automatic Gain Control] circuits or amplifiers

Abstract

A method for dynamic event detection based on content from a set of social networking systems including receiving content from the set of social networking systems, identifying a plurality of content associated with a geofence, the content that was generated within a predetermined time period, determining feature values from the plurality of content for each of a set of features, determining an event probability for the geofence based on the feature values, and detecting an event within the geofence in response to the event probability exceeding a threshold event probability.

Claims

1. A method for dynamic event extent determination, comprising:

receiving content from a set of social networking systems;

identifying a plurality of content, from the received content, associated with a geofence;

determining a plurality of feature values from the plurality of content;

reducing the plurality of feature values to a reduced plurality of feature values based on an event detection model;

using the event detection model, automatically detecting an event within a geofence of the plurality of geofences based on the reduced plurality of feature values;

in response to detecting the event within the geofence;

including the geofence in an event region associated with the event;

determining a geographic extent of the event region, comprising iteratively;

identifying a proximal geofence from among a plurality proximal geofences located proximal the geofence;

analyzing the proximal geofence for the event based on the reduced plurality of feature values; and

in response to detecting the event in the proximal geofence, including the proximal geofence in the event region;

detecting an event termination based on another set of feature values from among the plurality of feature values; and

associating the event region with the detected event.

2. The method of claim 1, further comprising determining a baseline value for a feature from historic content associated with each geofence of the plurality;

wherein automatically detecting the event comprises;

detecting the event in response to the feature value for the feature deviating from the baseline value for the geofence.

3. The method of claim 2, wherein the baseline value comprises a baseline feature frequency for the respective geofence, wherein detecting the event comprises detecting the event in response to the feature frequency exceeding the baseline feature frequency.

4. The method of claim 3, wherein the content comprises posts, wherein the baseline feature frequency comprises a baseline posting frequency for the respective geofence, wherein detecting the event comprises detecting the event in response to the posting frequency within the geofence exceeding the baseline posting frequency for the geofence.

5. The method of claim 3, wherein the baseline feature frequency comprises a baseline keyword frequency for the respective geofence, wherein detecting the event comprises detecting the event in response to a keyword frequency within the geofence exceeding the baseline keyword frequency for the geofence by a predetermined amount.

6. The method of claim 5, wherein the predetermined amount for the proximal geofence is determined based on a size of the event region.

7. The method of claim 1, wherein each geofence of the plurality of geofences represents a different physical region, each physical region having a common unit area.

8. The method of claim 1, wherein the proximal geofence is contiguous with the geofence within the event region.

9. The method of claim 1,

wherein reducing the plurality of feature values comprises aggregating the selected content into a content feed;

and

further comprising sending the content feed, in association with an identifier for the event region, to a remote user device.

10. The method of claim 9, wherein aggregating the selected content into a content feed is repeatedly performed as new content is received from the set of social networking systems;

and

in response to detecting the event termination, ceasing content selection associated with the event.

11. The method of claim 10, wherein detecting the event termination comprises:

determining an event probability from the second set of feature values; and

detecting the event termination in response to determination of the event probability falling below a threshold event probability.

12. The method of claim 11, further comprising maintaining the event probability for a predetermined period of time after determination that the event probability fell below the threshold event probability.

13. A method for dynamic event extent determination, comprising:

for each of the plurality of geofences, determining a baseline value for a feature from historic content associated with each geofence of the plurality;

receiving content from a set of social networking systems;

for each of a plurality of geofences;

identifying a plurality of content associated with the geofence; and

determining a feature value for the feature from the plurality of content;

detecting a first event within a first geofence of the plurality in response to the feature value deviating from the baseline value for the first geofence;

within a predetermined time window of detecting the first event, detecting a second event within a second geofence of the plurality in response to the feature value deviating from the baseline value for the second geofence;

determining that the first and second event are a common event;

determining a third geofence encompassing the first and second geofences; and

associating the third geofence with the common event;

selecting content associated with the common event from the plurality of content;

aggregating the selected content into a content feed;

sending the content feed, in association with an identifier for the event region, to a remote user device in response to receipt of a request from the remote user device;

receiving new content from the set of social networking systems;

after detecting the first and second events within the first and second geofences, respectively;

determining a second set of feature values for the third geofence from the new content;

detecting an event termination based on the second set of feature values; and

ceasing content selection associated with the event.

14. The method of claim 13, wherein the baseline value comprises a baseline keyword frequency for the respective geofence, wherein detecting the event comprises detecting the event in response to a keyword frequency exceeding the baseline keyword frequency by a predetermined amount.

15. The method of claim 13, wherein the content comprises posts, wherein the baseline value comprises a baseline posting frequency for the respective geofence, wherein detecting the event comprises detecting the event in response to the posting frequency within the geofence exceeding the baseline posting frequency for the geofence.

16. The method of claim 13, further comprising categorizing the first and second events with event categories, based on the respective feature values.

17. The method of claim 16, wherein determining that the first and second event are a common event comprises determining that the first and second events are categorized with a common event category.

18. The method of claim 13, wherein detecting the event termination comprises:

determining an event probability from the second set of feature values; and

detecting the event termination in response to determination of the event probability falling below a threshold event probability.

19. The method of claim 18, further comprising maintaining the event probability for a predetermined period of time after determination that the event probability fell below the threshold event probability.