Event prediction using temporal and geospatial precursor networks

• US 9,727,822 B1
• Filed: 04/27/2015
• Issued: 08/08/2017
• Est. Priority Date: 05/09/2012
• Status: Active Grant

• Current Assignee: Priority 5 Holdings, Inc.
• Original Assignee: Priority 5 Holdings, Inc.
• Inventors: Dudenhoeffer, Donald D., Miller, Charles Q., Bierbaum, Allen D., McDermott, Anthony J., Miller-Klugman, David M.
• Primary Examiner(s): Wong, Lut
• Application Number:
  US14/696,962
• Time in Patent Office: 834 Days
• Field of Search: 706 47
• 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 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 50/265 Personal security, identity or safety
• CPC Class Codes: G08B 31/00 Predictive alarm systems characterised by extrapolation or other computation using updated historic data

Abstract

The present invention provides a system and method for providing an alert notification prior to occurrence of a consequent event. The present invention provides a decision maker with the means to reconstitute his unfiltered operational environment such that the information needed to make an informed decision is extracted from a vast array of data available and presented in a manner that allows the decision maker to focus on the aspects of the decision that are most important to arriving at the best course of action under the exigent circumstances that are present when a response to an unexpected and possibly deleterious event is required.

Claims

1. A method for providing an alert notification prior to occurrence of a consequent event, comprising:

providing, in a non-volatile computer memory, a precursor activity network comprising data associated with a plurality of precursor activities, each of the precursor activities having a weighting factor associated with, the precursor activity network being associated with a consequent event;

upon occurrence of an actual event, determining whether the actual event is consistent with at least one of the precursor activities in the precursor activity network, in accordance with criteria associated with each of the precursor activities;

if the actual event is consistent with at least one of the precursor activities in the precursor activity network, updating the precursor activity network by associating said one of the precursor activities with a geospatial location of the actual event;

calculating, after updating, an indicator value of the precursor activity;

if the indicator value exceeds a predefined threshold value, issuing an alert notification indicating an increased likelihood for the consequent event; and

displaying, on a display panel, the alert notification to an operator.

2. The method of claim 1, wherein criteria associated with the plurality of precursor activities comprises a time period and a geospatial location of each of the precursor activities.

3. The method of claim 2, further comprising:

if the time period associated with the actual event no longer satisfies a predefined time period associated with the precursor activity network, updating the precursor activity network by removing the actual event from the precursor activity network.

4. The method of claim 3, wherein at least one precursor activity in the precursor activity network is associated with a number counter which may be used to count a number of active qualifying actual events.

5. The method of claim 4, wherein calculating an indicator value comprises summing weighting factors of all triggered precursor activities.