Telephonic communications – With usage measurement – Call charge metering or monitoring
Patent
1998-02-09
1999-05-25
Loomis, Paul
Telephonic communications
With usage measurement
Call charge metering or monitoring
379134, 379189, 379198, H04M 1500
Patent
active
059076020
DESCRIPTION:
BRIEF SUMMARY
This invention relates to methods of, and systems for, detecting possible fraudulent use of a communications network.
Early detection of possible fraudulent use of communications services provided on a network is desirable in order to limit the financial liability of the network operator, the service provider and the customer, for convenience any or all of these will be referred to hereinafter as a protected person (PP).
A fraud detection system is known from WO-A-94 11959 (Coral Systems, Inc.) in which a parameter of a particular subscriber's current usage is compared with a subscriber-specific pattern of the particular subscriber's historic usage. For example, the subscriber-specific pattern can be a five day rolling average of the daily number of calls made by the subscriber, or it can be the maximum number of calls ever made by the subscriber in a day.
A fraud detection system is known from EP A 0 618 713 (AT&T Corporation) for providing fraud monitoring for a network customer for whom a telephone service provider (network operator) provides private network facilities via a Software Defined Network, access to which is obtained through the use of an authorisation code or toll-free telephone number ("800" number). In this system, a fraud monitoring processor uses the call attributes that are stored in a call detail record to derive various indices (e.g. average call duration for calls made using a selected authorisation code). The fraud monitoring processor then compares the call attributes and/or derived indices with preselected thresholds which the network customer (perhaps with the assistance of the telephone service provider) has previously selected or established as being indicative of abnormal network usage. The thresholds are established by the network customer in view of its particular needs or preferences.
According to a first aspect of the present invention, there is provided a method of detecting possible fraudulent use of a communications network, which network is arranged to provide call event records having fields containing call event data, at least one of the fields being designated a key field and containing a key relating to an entity as herein defined, the method comprising the steps of: obtained key (referred to as the "first combination") to access a first combination-associated profile set comprising be performed upon call event data to obtain a process result, by the process result, and at least one respective threshold for the or each said first combination-associated first store, a said first combination-associated first store reaches a threshold, at least one said predetermined outcome comprising a fraud risk indicator, fraud risk indicator, second store, and contents of said first combination-associated second store reaches a threshold; said first combination-associated profile set, thereby first store or stores by the respective process result, and, if a threshold is reached, fraud risk indicator is generated, risk indicator, and, if a threshold is reached,
In this first aspect, preferably a second of the fields of the call event record is designated a key field and contains a key relating to an entity as herein defined, and the method comprises the further steps of: key; obtained key (referred to as the "second combination") to access a respective second combination-associated profile set comprising, be performed upon call event data of a call event record to obtain a process result, incremented by the process result, and at least one respective threshold for the or each said second combination-associated first store, a said second combination associated first store reaches a threshold, at least one said predetermined outcome comprising a fraud risk indicator, event record being currently processed to be incremented by a fraud risk indicator generated by a profile of its own profile set; said respective second combination-associated profile set, thereby first store or stores by the respective process result, and, if a threshold is reached, fraud risk indicator is generated, r
REFERENCES:
patent: 4893330 (1990-01-01), Franco
patent: 5351287 (1994-09-01), Bhattacharyya et al.
patent: 5603906 (1997-02-01), Phelps
patent: 5706338 (1998-01-01), Relyea et al.
patent: 5805686 (1998-09-01), Moller et al.
Eaden Russel
Gibbins Christopher John
Peel Stephen John
Regnault John Christopher
Scott Michael
British Telecommunications public limited company
Loomis Paul
LandOfFree
Detecting possible fraudulent communication usage does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Detecting possible fraudulent communication usage, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Detecting possible fraudulent communication usage will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-406132