System for detection and prevention of telecommunications fraud prior to call connection
First Claim
1. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said method comprising, in combination:
- receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree;
applying said first fraud screening tree to said attempted call based on said information carried in said data message, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number.
1 Assignment
0 Petitions
Accused Products
Abstract
An improved system for detecting and preventing telecommunications fraud prior to call connection, in which attempted calls are screened for fraud at the point of calling card validation in the SCP. The SCP accesses a database holding a variety of fraud decision nodes that may be interrelated and grouped together to cooperatively define fraud screening trees against which attempted calls may be tested. A fraud screening tree ultimately produces a fraud prediction based on a variety of information concerning the attempted call. The present invention therefore facilitates fraud screening with greater granularity and customization.
-
Citations
81 Claims
-
1. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree;
applying said first fraud screening tree to said attempted call based on said information carried in said data message, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14)
-
-
15. In a telecommunications network, a method of determining that an attempted call is likely to be fraudulent and responsively preventing said attempted call from being connected to a dialed number, said attempted call defining a customer billing number, a call origination number and said dialed number, said telecommunications network including a querying system and a service control point (SCP), said querying system receiving an indication of said attempted call and responsively passing a data message to said SCP, said data message defining information about said attempted call, said method comprising, in combination:
-
storing, in a database accessible by said SCP, a plurality of fraud decision nodes, wherein each of said decision nodes defines criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, said fraud decision node points to a next one of said fraud decision nodes, a given group of said fraud decision nodes thereby cooperatively defining a fraud screening tree through which said attempted call may be tested, said fraud screening tree including a final fraud decision node that defines a fraud prediction for said attempted call;
receiving said data message at said SCP;
verifying said customer billing number;
selecting, based on said customer billing number, a first fraud decision node of said fraud screening tree;
applying said fraud screening tree to said attempted call;
based on said fraud prediction, generating a response data message indicating whether or not to block said attempted call from being connected to said dialed number;
passing said response data message to said querying system; and
if said response data message indicates to block said attempted call from being connected, preventing said attempted call from being connected to said dialed number. - View Dependent Claims (16, 17, 18, 19, 20, 21, 22, 23, 24)
-
-
25. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said information including a customer billing number, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes defining criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, said fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree;
selecting, based on said customer billing number, the starting node of said first fraud screening tree;
starting with the starting node, applying the plurality of fraud decision nodes defining the first fraud screening tree, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number. - View Dependent Claims (26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36)
-
-
37. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes defining criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, each of a plurality of said fraud decision nodes points to a next one of said fraud decision nodes, a first subset of said fraud decision nodes defining a fraud screening tree, said fraud screening tree including a starting node and a final fraud decision node that determines a fraud prediction for said attempted call;
applying the starting node of said fraud screening tree to said attempted call based on said information carried in said data message, and applying the next one of said fraud decision nodes pointed to by the starting node based on satisfaction of the starting node criteria;
based on satisfaction of the criteria in each applied fraud decision node, applying the next one of each fraud decision nodes pointed to by each applied decision node until the final fraud decision node determines the fraud prediction;
based on said fraud prediction, generating a response message indicating whether or not to block said attempted call from being connected to said dialed number;
passing said response message to said querying system; and
if said response message indicates to block said attempted call from being connected, preventing said attempted call from being connected to said dialed number. - View Dependent Claims (38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48)
-
-
49. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said information including a customer billing number associated with a predetermined level of fraud screening, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes defining criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, said fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree;
selecting, based on said predetermined level of fraud screening, the starting node of said first fraud screening tree;
starting with the starting node, applying the plurality of fraud decision nodes defining the first fraud screening tree, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number. - View Dependent Claims (50, 51, 52, 53, 54, 55, 56, 57, 58, 59)
-
-
60. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said information including a customer billing number said billing number associated with a market class, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes defining criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, said fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree;
selecting the starting node of said first fraud screening tree based on said market class;
starting with the starting node, applying the plurality of fraud decision nodes defining the first fraud screening tree, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number. - View Dependent Claims (61, 62, 63, 64, 65, 66, 67, 68, 69)
-
-
70. In a telecommunications network, a method for determining whether to block an attempted call from being connected to a dialed number due to a likelihood that the attempted call is fraudulent, said attempted call being defined by information carried in a data message provided by a querying system, said information including a customer billing number, said method comprising, in combination:
-
receiving said data message into a computer system, said computer system including a database holding a plurality of fraud decision nodes defining criteria against which said attempted call may be tested and wherein, based on satisfaction of said criteria, said fraud decision nodes cooperatively defining fraud screening trees each having a starting node, a first subset of said fraud decision nodes cooperatively defining a first fraud screening tree, each of the plurality of said fraud decision nodes comprising a database table, wherein a row in each fraud decision node defines a next fraud decision node based on specified criteria;
selecting, based on said customer billing number, the starting node of said first fraud screening tree;
starting with the starting node, applying the plurality of fraud decision nodes defining the first fraud screening tree, said first fraud screening tree producing a determination of whether or not to block said attempted call due to a likelihood that the attempted call is fraudulent; and
before said attempted call is connected to said dialed number, returning to said querying system a response data message indicative of said determination, whereby, if said determination is to block said attempted call, said attempted call is not connected to said dialed number. - View Dependent Claims (71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81)
-
Specification