Patent Application
20170169485
This application claims the benefit of Singapore Patent Application No. 10201510132U filed Dec. 10, 2015, which is hereby incorporated by reference in its entirety.
The present disclosure relates to methods and systems for identifying individuals who are liable to make a donation to a charitable organization (a “charity”), so that those individuals can be contacted to solicit a donation.
There are many factors which influence whether an individual gives to a charity, and if so how large a donation. One is the financial situation of the individual, and what he or she can afford. Another is the character of the individual, and how generous he or she is.
Even for individuals who have disposable income, and a tendency to donate a proportion of it to a worthwhile cause, different individuals may be more or less likely to make a donation to a given charity. For example, this depends upon the stated objective of the charity. Certain individuals, for example, are more likely to make a donation to a charity helping animals. Other individuals are more likely to make a donation to a charity which assists people with immediate needs (e.g. victims of a natural disaster). Yet further individuals are more likely to donate money to a charity with a less immediate objective, such as one which conducts medical research with the hope of discovering new medical treatments for use many years in the future.
Furthermore, individuals react differently to different advertising campaigns. A first individual may be moved strongly to an advertising campaign using images depicting victims of a natural disaster, while a second individual may be repelled by such images and more strongly moved to make a charitable donation by more positive images, such as images used by an arts charity and depicting a theatrical production which could be paid for by charitable donations.
Many charitable organizations devote a significant proportion of their income to advertising campaigns, and part of this budget will be wasted if it is used to advertise to individuals who are not able or willing to make a donation to anyone, who are not in sympathy with the objectives of the charitable organization, or who are unmoved by the images and sounds used in the advertising campaign. By improving the targeting of the advertising, the revenue of the charities can be improved, to the general benefit of society as a whole.
In general terms, the present disclosure proposes methods and systems for identifying a subset (“segment”) of a population of individuals for a charitable organization to target in an advertising campaign, based on transactional data describing payment transactions made by some or all of the population of individuals and demographic and/or location data relating to the individuals.
Specifically, the disclosure proposes using a database of payment transactions made by the population of individuals and a database of demographic and/or location data for the corresponding individuals, to develop a predictive model for predicting the likelihood that a candidate individual in the population will make a charitable donation, the predictive model being a function of data values describing the history of the payment transactions and/or demographic and/or location data for the candidate individual.
Once the model is developed, the predictive model is used to identify a segment of the population of individuals for whom, according to the model, the likelihood of making a charitable donation is high, and then individuals in that segment of the population are solicited for donations.
The database of payment transactions includes data describing past payment transactions made to a charitable organization. Such transaction data is particularly useful for identifying candidate individuals who have previously made a donation to a charitable organization, and so are more likely to do so in the future. However, a useful predictive model may be developed even in the case of a candidate individual who, according to the payment transaction database, has not made a donation to a charitable organization.
The predictive model may be framed as a decision tree, by which a predictive value may for the candidate individual may be obtained by selecting a path through the decision tree according to the data values describing the history of the payment transactions and demographic and/or location data for the candidate individual.
The term “payment transaction” is used to refer to an automated process in which a payment is made to an entity, such as using a payment card. The term “payment card” refers to any suitable cashless payment device, such as a credit card, a debit card, a prepaid card, a charge card, a membership card, a promotional card, a frequent flyer card, an identification card, a prepaid card, a gift card, and/or any other device that may hold payment account information, such as mobile phones, Smartphones, personal digital assistants (PDAs), key fobs, transponder devices, NFC-enabled devices, and/or computers.
The term “charitable organization” (charity) is used to mean an organization which has as its primary objective a non-profit activity. In some countries charitable organizations are granted a specific legal status, and if so the definition of the term charitable organization in such countries may be depend at least partly upon this legal status.
Embodiments of the disclosure will now be described for the sake of non-limiting example only, with reference to the following drawings in which:
Schematically, the computer system includes a processing unit 10 with access to four types of database. First, there is a database 20 describing payment transactions made by a plurality of individuals. The database 20 may for example be obtained from a payment network, such as the one operated by MasterCard International Incorporated, and relate to payment transactions made by payment cards.
Secondly, the processing unit 10 has access to a database 30 which contains, in respect of at least some of the population of individuals, contains demographic data and/or location data. The demographic data may include any one of more of: gender, age, and/or marital status. The location data may for example be zipcode (postcode) for the corresponding individuals.
Thirdly, the processing unit 10 has access to a database 40 containing advertising information describing messages which a charity wishes to transmit to appropriate individuals of the population, as part of an advertising campaign.
Fourthly, the processing unit 10 has access to a database 50 of contact information for the individuals, such as a corresponding email address, postal address or telephone number for each of the individuals.
The technical architecture 220 includes a processor 222 (which may be referred to as a central processor unit or CPU, and which plays the role of the processing unit 10 in the schematic description given above). The processor 222 in communication with memory devices including secondary storage 224 (such as disk drives), read only memory (ROM) 226, and random access memory (RAM) 228. The databases 20, 30, 40 and 50 may be stored on any one or more of these memory devices.
The processor 222 may be implemented as one or more CPU chips. The technical architecture 220 may further include input/output (I/O) devices 230, and network connectivity devices 232.
The secondary storage 224 typically includes one or more disk drives or tape drives and is used for non-volatile storage of data and as an over-flow data storage device if RAM 228 is not large enough to hold all working data. Secondary storage 224 may be used to store programs which are loaded into RAM 228 when such programs are selected for execution. In this embodiment, the secondary storage 224 has a mobile wallet registration component 224a, and a mobile wallet payment authorization component 224b including non-transitory instructions operative by the processor 222 to perform various operations of the method of the present disclosure. The ROM 226 is used to store instructions and perhaps data which are read during program execution. The secondary storage 224, the RAM 228, and/or the ROM 226 may be referred to in some contexts as computer readable storage media and/or non-transitory computer readable media.
I/O devices 230 may include printers, video monitors, liquid crystal displays (LCDs), plasma displays, touch screen displays, keyboards, keypads, switches, dials, mice, track balls, voice recognizers, card readers, paper tape readers, or other well-known input devices.
The network connectivity devices 232 may take the form of modems, modem banks, Ethernet cards, universal serial bus (USB) interface cards, serial interfaces, token ring cards, fiber distributed data interface (FDDI) cards, wireless local area network (WLAN) cards, radio transceiver cards that promote radio communications using protocols such as code division multiple access (CDMA), global system for mobile communications (GSM), long-term evolution (LTE), worldwide interoperability for microwave access (WiMAX), near field communications (NFC), radio frequency identity (RFID), and/or other air interface protocol radio transceiver cards, and other well-known network devices. These network connectivity devices 232 may enable the processor 222 to communicate with the Internet or one or more intranets. With such a network connection, it is contemplated that the processor 222 might receive information from the network, or might output information to the network in the course of performing the above-described method operations. Such information, which is often represented as a sequence of instructions to be executed using processor 222, may be received from and outputted to the network, for example, in the form of a computer data signal embodied in a carrier wave.
The processor 222 executes instructions, codes, computer programs, scripts which it accesses from hard disk, floppy disk, optical disk (these various disk based systems may all be considered secondary storage 224), flash drive, ROM 226, RAM 228, or the network connectivity devices 232. While only one processor 222 is shown, multiple processors may be present. Thus, while instructions may be discussed as executed by a processor, the instructions may be executed simultaneously, serially, or otherwise executed by one or multiple processors.
Although the technical architecture 220 is described with reference to a computer, it should be appreciated that the technical architecture may be formed by two or more computers in communication with each other that collaborate to perform a task. For example, but not by way of limitation, an application may be partitioned in such a way as to permit concurrent and/or parallel processing of the instructions of the application. Alternatively, the data processed by the application may be partitioned in such a way as to permit concurrent and/or parallel processing of different portions of a data set by the two or more computers. In an embodiment, virtualization software may be employed by the technical architecture 220 to provide the functionality of a number of servers that is not directly bound to the number of computers in the technical architecture 220. In an embodiment, the functionality disclosed above may be provided by executing the application and/or applications in a cloud computing environment. Cloud computing may include providing computing services via a network connection using dynamically scalable computing resources. A cloud computing environment may be established by an enterprise and/or may be hired on an as-needed basis from a third party provider.
It is understood that by programming and/or loading executable instructions onto the technical architecture 220, at least one of the CPU 222, the RAM 228, and the ROM 226 are changed, transforming the technical architecture 220 in part into a specific purpose machine or apparatus having the novel functionality taught by the present disclosure. It is fundamental to the electrical engineering and software engineering arts that functionality that can be implemented by loading executable software into a computer can be converted to a hardware implementation by well-known design rules.
Various operations of the methods carried out by the computer system 10 will now be described with reference to
In a first step 301 of the method of
In step 303, the processing unit 10 identifies a subset of the total population of individuals (the “training subset”) for whom reliable data exists in both the databases 10, 20. For example, the processing unit 10 may check that for a given individual the database 20 contains records of a sufficiently large number of payment transactions to be statistically typical of the individual's total payment behavior (for example, the number of payment transactions (e.g. within a predefined time window) is above a predefined threshold).
In step 304, the processing unit 10 searches the transaction data to identify individuals who have made a payment to any of a predefined set of charitable organization (which may for example by all the charitable organizations operating in the jurisdiction in which the population of individuals live). Thus, it forms a number of records corresponding to the respective training subset of individuals. Each record includes a respective flag value indicating whether the respective individual has made a payment to one of the set of charitable organizations, and a respective set of descriptor values based on the data from the databases 20 and/or 30 describing the respective individual. Thus, the descriptor values may describe the previous payment transactions of the individual (for example, the number of previous payment transactions (e.g. during a certain time window), the total value of those transactions, the median value of the transactions, etc.) and/or one or more demographic characteristic(s) of the individual and/or a geographical location associated with the individual (e.g. his or her billing address). The geographical location may for example be expressed as a zipcode, or converted into another format, such as a variable indicating that the zipcode represents a location with certain pre-defined characteristics (e.g. it is a location in the city or in countryside, or it is a region statistically associated with a certain wealth level, e.g. a place where affluent individuals tend to live).
In step 305, the processing unit 10 generates a predictive model using the records about the training subset of individuals as training data. The predictive model attempts to predict the flag value from the descriptor values. The predictive model is typically an adaptive model, and typically generated iteratively. Conveniently, the predictive model may be a decision tree, of the kind shown in
For example, in the case of an individual whose payment transactions in the past month total $12,500, who is female and aged 65, the decision tree reaches position E, which is associated with a certain predictive value (e.g. 65%) that the individual has made a charitable donation. The predictive value has been found by observing that 65% of the training subset of individuals for whom the questions had the same answers to the demographic/location questions (i.e. 65% of the individuals in the training subset who were women below the age of 70 who had payment transactions totalling over S$10,000 in the past month) had made a charitable donation according to the database 10. Conversely in the case of an individual whose payment transactions totalled S$9,500, is aged 75 and has a zipcode in an area which has previously been registered as being in the affluent, the path through the decision tree reaches position F, which is associated with a different predictive value, such as 80%. An individual for whom the path reaches the position F is thus more likely to engage in charitable giving than an individual for whom the path reaches position E.
The decision tree of
A large number of questions can be used. For example, although question 4 determines whether the individual has an age above a threshold value of 40, any other age may be used as the threshold value. The questions used in the decision tree are chosen to give maximum discrimination (i.e. predictive value) for the flag value. The number of questions may be higher or lower than the 7 shown in
A number of automatic algorithms exist for constructing a decision tree. Many such algorithms are iterative. Some such algorithms are described in Rokach, Lior; Maimon, O. (2008) “Data mining with decision trees: theory and applications”, World Scientific Pub Co Inc. (see also Chapter 1 Barry de Ville and Padriac Neville (2013) “Decision Trees for Analytics Using SAS Enterprise Miner”, SAS Institute Inc.). The most commonly used algorithm is called “top-down induction of decision trees” (TDIDT).
In step 505, a determination is made of whether a termination criterion has been met. For example, the termination criterion may be whether steps 501-504 have been carried out for all candidate individuals for whom data exists in databases 30 and 50. Alternatively, if a charitable organization is limited in the number of advertising messages which can be sent, the termination criterion may be whether this number of advertising messages has been sent. If step 505 determines that the termination criterion is met, the method terminates. Otherwise, the method returns to step 501 in which a new candidate individual is selected (one for whom the method of
Many variations of the present scheme are possible. For example, a noted above certain individuals are more likely to contribute to a certain class of charity (e.g. an animal charity). Thus, when the advertising campaign for which data is stored in the database 40 is for a charity in a certain class (e.g. an animal charity), the determination made in step 304 may relate only to charities of the same class (i.e. step 304 determines whether the individual has previously made a donation to an animal charity). The class of charity may be defined by one or more charitable criteria, e.g. one of the charitable criteria may be whether the beneficiaries of the charity are animal or human, another of the charitable criteria may be whether object of the charity is to improve the health of the beneficiaries, yet another may be the type of images the charitable organization uses in advertising messages, e.g. shocking images or positive ones.
Furthermore, the predictive value for a given candidate individual obtained at step 502 may take further information into account than the result of the decision tree of
In this case, the processing unit 10 first determines in step 601 whether payment transaction history of the candidate individual meets one or more payment transaction criteria, e.g. ones which are not used in the decision tree. For example, one of the criteria may be whether the candidate individual has made a donation to any charity, or to a charity in the same class as the charity which the method of
In step 602 the decision tree is followed to obtain a predictive value for the candidate.
In step 603, the predictive value obtained using the decision tree is modified based on the payment transaction metric value obtained in step 601. For example, let us consider the case that there is only one payment transaction criterion, which is whether the candidate individual has previously made a donation of the specified type. If step 601 concluded that the candidate individual has done this, then the predictive value obtained in step 603 may be modified by making it closer to 100%, e.g. by increasing it such that the difference between it and 100% is halved. Conversely, if step 601 concluded that the candidate individual has not previously made a donation of the specified type, then the predictive value obtained in step 602 may be reduced, e.g. by dividing it by two.
Whilst the foregoing description has described exemplary embodiments, it will be understood by those skilled in the art that many variations of the embodiment can be made within the scope and spirit of the present disclosure.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10201510132U | Dec 2015 | SG | national |
