TWM605348U - Financial Commodity Recommendation System - Google Patents

Abstract

A financial product recommendation system includes a server unit and terminal electronic devices. The server unit stores multiple personal data related to multiple reference customers, and generates multiple personal characteristic data related to the reference customers based on the personal data, each of which includes A plurality of personal characteristic values, and based on the personal characteristic data, a predetermined clustering algorithm is used to generate a plurality of cluster definition data. The terminal electronic device transmits a product suggestion request containing a to-be-analyzed personal data related to a to-be-analyzed customer to the server unit. The server unit generates the to-be-analyzed personal characteristic data and the cluster definition data based on the to-be-analyzed personal data, and generates a clustering result related to one of the cluster definition data.

Description

Financial Commodity Recommendation System

This model relates to a recommendation system, especially a financial product recommendation system.

According to statistics, nearly 60% of the people do not know how to plan financially, including not knowing the pension system, not knowing the gap in their pension, and not knowing where to start planning. Therefore, it is possible that the lack of knowledge of retirement financial management may cause miscalculation of savings goals and future needs after retirement.

At present, if the public wants to learn more about the above-mentioned financial planning, they usually have to go to the bank to ask the financial specialist in person, and time-consuming problems may arise.

Therefore, the purpose of the present invention is to provide a financial product recommendation system that can improve at least one of the disadvantages of the prior art.

Therefore, the financial product recommendation system provided by the present invention includes a server unit and a terminal electronic device capable of communicating with the server unit via a communication network. The server unit stores a plurality of personal data respectively related to a plurality of reference customers, and a plurality of financial product holding data respectively related to the reference customers. The server unit generates a plurality of personal characteristic data related to the reference customers according to the personal data, and each of the personal characteristic data includes a plurality of personal characteristic values; the server unit generates a plurality of personal characteristic data according to the personal characteristic data , Using a predetermined clustering algorithm to generate a plurality of cluster definition data, each cluster definition data includes a plurality of characteristic value ranges respectively related to the personal characteristic values; the terminal electronic device transmits a data that includes a client to be analyzed A product suggestion request of the personal data to be analyzed is sent to the server unit; the server unit generates a personal characteristic data to be analyzed based on the personal data to be analyzed, and the personal characteristic data to be analyzed includes multiple definitions respectively related to the clusters The individual characteristic values to be analyzed in the range of the characteristic values of the data; the server unit generates a clustering result related to one of the cluster definition data based on the waiting individual characteristic values and the cluster definition data, wherein, The characteristic value of the individual waiting to be analyzed conforms to the characteristic value range of the cluster definition data related to the grouping result; the server unit sends a suggested combination of financial products corresponding to the cluster definition data related to the grouping result to the Terminal electronic device.

In some implementations, the server unit also uses a non-linear regression method based on the cluster definition data to establish a plurality of income estimation models corresponding to the cluster definition data, and the clustering result is based on the clustering result. The income estimation model corresponding to the related cluster definition data is set as a target income estimation model; the server unit estimates an estimated personal income of the client to be analyzed based on the personal data to be analyzed and using the target income estimation model ; The server unit generates a financial product investment amount corresponding to the recommended combination of financial products based on the estimated personal income.

In some implementation aspects, each personal data includes a geographic environment data related to a geographic location of the reference customer, and a public government data related to the reference customer; the personal data to be analyzed includes information related to the to-be-analyzed A to-be-analyzed geographical environment data of a customer's geographical location to be analyzed and a to-be-analyzed government data related to the to-be-analyzed customer.

In some implementation aspects, the geographic environment information is related to at least one of land administration status, public facility status, and private facility status; and the government information is related to average remaining life of multiple ages, income and expenditure of multiple administrative regions, At least one of and employment status.

In some implementation aspects, the government data to be analyzed included in the personal data to be analyzed is downloaded from a government database.

In some implementation aspects, the predetermined grouping algorithm is a k-means algorithm.

The effect of the present invention is that the financial product recommendation system of the present invention uses the server unit to generate the cluster definition data based on the personal characteristic data of the reference customers and using the k-average algorithm, so as to receive data from When the terminal electronic device requests the product suggestion, the grouping result can be generated based on the cluster definition data and the personal data to be analyzed included in the product suggestion request, and then the corresponding financial product suggestion can be based on the grouping result The combination is transmitted to the terminal electronic device, so it can indeed achieve the purpose of new cost.

Referring to FIG. 1, an embodiment of the financial product recommendation system of the present invention includes a server unit 10 and a terminal electronic device 20 capable of communicating with the server unit 10 via a communication network 200 (for example, the Internet) .

The terminal electronic device can be a portable electronic device (for example, a smart phone, a tablet computer, a notebook computer, etc.) or a personal computer, but is not limited to the above.

In this embodiment, the server unit 10 is, for example, a computer host belonging to a bank, and is suitable for connecting to a government database 300, and stores a plurality of personal data respectively related to a plurality of reference customers. And a number of financial product holding data related to these reference customers.

Each personal data includes the corresponding basic data of a person related to the reference customer, a geographic environment data related to a geographic location of the reference customer, and a public government information related to the reference customer. The basic personal information can be the reference customer’s gender, age, occupation, job title, income... etc. The geographic location can be the reference customer’s household registration, and the geographic environment information is related to the land administration, public facilities and private facilities At least one of the conditions, and for example, how many hospitals are within a preset distance from the geographic location, the distance between the geographic location and the nearest MRT station, or how many are within a preset distance from the geographic location Disgusting facilities...etc. The government information is related to the income and expenditure of multiple administrative regions, and the employment status (employment rate). The above is only an example, and for example, the average income per person and average per person of the administrative area where the geographical location is located Expenditure, employment rate of the administrative region where the geographic location is located, etc., are not limited to the above.

1 and FIG. 2, the following describes the steps of performing a one-grouping procedure in this embodiment.

First, in step S11, the server unit 10 generates a plurality of personal characteristic data respectively related to the reference customers according to the personal data, and each of the personal characteristic data includes a plurality of personal characteristic values.

More specifically, the personal characteristic values contained in each personal characteristic data, such as age, annual income, distance between the residence and the nearest MRT station, the number of hospitals within a preset distance from the residence, and distance The number of disgusting facilities within a preset distance from the household registration place...etc.

Next, in step S12, the server unit 10 uses a predetermined clustering algorithm to generate a plurality of cluster definition data based on the personal characteristic data, and each cluster definition data includes a plurality of cluster definition data respectively related to the personal characteristic values. Characteristic value range.

In this embodiment, the predetermined clustering algorithm is k-means clustering, and each of the cluster definition data generated by the server unit 10 includes the characteristic value ranges The number of hospitals whose annual income is between 1.5 million and 2 million, the distance between the household registration and the nearest MRT station is between 300 and 500 meters, and the number of hospitals within a preset distance from the household registration is between 2 and 5... Etc., but not limited to the above. Furthermore, after the server unit 10 generates the cluster definition data, it then creates a plurality of correspondences respectively indicating a plurality of suggested combinations of financial products and the cluster definition data. The suggested combinations of financial products are generated by experts for the cluster definition data after the server unit 10 generates the cluster definition data.

1 and 3, the following describes the steps of a financial product recommendation program executed by this embodiment.

First, in step S21, the server unit 10 receives a product suggestion request from the terminal electronic device 20 related to a to-be-analyzed personal data of a to-be-analyzed customer.

More specifically, the to-be-analyzed personal data includes a to-be-analyzed geographic environment data related to a to-be-analyzed geographic location of the to-be-analyzed customer and a to-be-analyzed government data that is publicly related to the to-be-analyzed customer. As mentioned above, the geographic location to be analyzed may be the residence address of the customer to be analyzed, and the geographic environment data to be analyzed related to the geographic location to be analyzed may be between the residence address of the customer to be analyzed and the nearest MRT station The number of hospitals within a predetermined distance from the household registration address of the client to be analyzed, the number of disgusting facilities within a predetermined distance from the household registration address of the client to be analyzed, etc. The government information to be analyzed may be the average income of the administrative region related to the household registration address of the client to be analyzed, and is not limited to the above. In this embodiment, the government data to be analyzed included in the personal data to be analyzed is downloaded from the government database 300 by the server unit.

Next, in step S22, after the terminal electronic device 20 transmits the product suggestion request to the server unit 10, that is, after the server unit 10 receives the product suggestion request, the server unit 10 responds to the product suggestion request The to-be-analyzed personal data generates a to-be-analyzed personal characteristic data, and the to-be-analyzed personal characteristic data includes a plurality of to-be-analyzed personal characteristic values respectively related to the characteristic value ranges of the cluster definition data. Then, step S23 is executed.

In step S23, the server unit 10 generates a clustering result related to one of the cluster definition data according to the waiting analysis personal characteristic value and the cluster definition data. The range of the characteristic values of the cluster definition data related to the clustering result generated by the server unit 10 meets the characteristic value of the waiting individual.

The following table shows examples of cluster definition data and personal data to be analyzed. Cluster definition data A Number of factories within 3km of the household registration area 1~2 There is a large hospital within 5km of the household registration area 0 National income ranking of the administrative district to which the household registration place belongs Next 20~25% Cluster definition data B Number of factories within 3km of the household registration area 0 There is a large hospital within 5km of the household registration area 1~3 National income ranking of the administrative district to which the household registration place belongs Top 20~25% Personal data to be analyzed Number of factories within 3km of the household registration area 0 There is a large hospital within 5km of the household registration area 2 National income ranking of the administrative district to which the household registration place belongs Top 22%

According to the cluster definition data A and cluster definition data B shown above, since the personal data to be analyzed conforms to the characteristic value ranges included in the cluster definition data B, the clustering result generated by the server unit 10 is Will be related to the cluster definition data B.

Then, in step S24, the server unit 10 transmits the suggested combination of financial products corresponding to the cluster definition data related to the clustering result to the terminal electronic device 20. Following the previous example, the proposed combination of financial products corresponding to cluster definition data A contains large-value protection insurance, and the proposed combination of financial products corresponding to cluster definition data B contains annuity insurance.

1 and 4, the following describes the steps of an income estimation procedure executed by this embodiment.

In step S31, the server unit 10 uses a nonlinear regression method according to the cluster definition data to establish a plurality of income estimation models corresponding to the cluster definition data.

Then in step S32, the server unit 10 sets the income estimation model corresponding to the cluster definition data related to the grouping result as a target income estimation model according to the grouping result.

Following the previous example, the server unit 10 will set the revenue estimation model corresponding to the cluster definition data B as the target revenue estimation model. Then, step S33 is executed.

Referring to step S33, after the server unit 10 sets the target income estimation model, it estimates an estimated personal income of the client to be analyzed based on the personal data to be analyzed and the target income estimation model. Next, in step S34, a financial product investment amount corresponding to the proposed combination of financial products is generated based on the estimated personal income.

For example, the server unit 10 sets, for example, 1/3 of the estimated personal income as the financial product investment amount, and transmits the financial product recommended combination and the financial product investment amount to the terminal electronic device 20.

In summary, the new type of financial product recommendation system uses the server unit 10 to generate the cluster definition data based on the personal characteristic data of the reference customers and the k-average algorithm, so that it can receive data from When the terminal electronic device 20 requests the commodity suggestion, the grouping result can be generated based on the cluster definition data and the personal data to be analyzed contained in the commodity suggestion request, and then the corresponding financial commodity can be classified according to the grouping result The proposed combination is transmitted to the terminal electronic device 20, so it can indeed achieve the purpose of new cost.

However, the above-mentioned are only examples of the present model, and should not be used to limit the scope of implementation of the present model, all simple equivalent changes and modifications made in accordance with the patent scope of the present model application and the contents of the patent specification still belong to This new patent covers the scope.

10: Server unit 20: Terminal electronics 200: Communication network 300: Government Database S11~S12: steps S21~S24: steps S31~S34: steps

Other features and effects of the present invention will be clearly presented in the embodiments with reference to the drawings, in which: FIG. 1 is a schematic diagram of a hardware connection relationship of an embodiment of the new financial product recommendation system; Figure 2 is a flowchart of a grouping procedure executed by this embodiment; Figure 3 is a flowchart of a financial product recommendation program executed by this embodiment; and Fig. 4 is a flowchart of an income estimation procedure executed by this embodiment.

10: Server unit

20: Terminal electronics

200: Communication network

300: Government Database

Claims (6)

A financial product recommendation system, including: A server unit storing a plurality of personal data related to a plurality of reference customers, and a plurality of financial product holding data related to the reference customers; and A terminal electronic device that can communicate with the server unit via a communication network; Wherein, the server unit generates a plurality of personal characteristic data respectively related to the reference customers according to the personal data, and each of the personal characteristic data includes a plurality of personal characteristic values; The server unit uses a predetermined clustering algorithm to generate a plurality of cluster definition data according to the personal characteristic data, and each cluster definition data includes a plurality of characteristic value ranges respectively related to the personal characteristic values; The terminal electronic device transmits a product suggestion request containing a to-be-analyzed personal data related to a to-be-analyzed customer to the server unit; The server unit generates a to-be-analyzed personal characteristic data based on the to-be-analyzed personal data, and the to-be-analyzed personal characteristic data includes a plurality of to-be-analyzed personal characteristic values respectively related to the characteristic value ranges of the cluster definition data; The server unit generates a grouping result related to one of the cluster definition data according to the waiting analysis personal characteristic value and the cluster definition data, wherein the waiting analysis personal characteristic value matches the grouping result related to the grouping result. The range of these characteristic values of the cluster definition data; The server unit sends a suggested combination of financial products corresponding to the cluster definition data related to the clustering result to the terminal electronic device. The financial product recommendation system according to claim 1, wherein the server unit also uses a nonlinear regression method to establish a plurality of income estimation models corresponding to the cluster definition data according to the cluster definition data, and according to The grouping result sets the income estimation model corresponding to the cluster definition data related to the grouping result as a target income estimation model; The server unit estimates an estimated personal income of the client to be analyzed based on the personal data to be analyzed and using the target income estimation model; and The server unit generates a financial product investment amount corresponding to the recommended combination of financial products based on the estimated personal income. The financial product recommendation system according to claim 1, wherein each personal data includes a geographic environment data related to a geographic location of the reference customer, and a public government data related to the reference customer; The to-be-analyzed personal data includes a to-be-analyzed geographic environment data related to a to-be-analyzed geographic location of the to-be-analyzed customer and a to-be-analyzed government data that is related to the to-be-analyzed customer and is public. The financial product recommendation system according to claim 3, wherein the geographic environment data is related to at least one of land administration status, public facility status, and private facility status; and The government information is related to at least one of the income and expenditure of multiple administrative regions, and employment status. The financial product recommendation system according to claim 3, wherein the government data to be analyzed included in the personal data to be analyzed is downloaded from a government database. The financial product recommendation system according to claim 1, wherein the predetermined grouping algorithm is a k-average algorithm.

Images