KR20220022167A - Apparatus and metho for recommendation of financial products based on aritificail intelligence using unstrucred data - Google Patents

Abstract

According to one technical aspect of the present invention, an artificial intelligence (AI)-based financial product recommendation apparatus using unstructured data comprises: an unstructured data collection unit collecting unstructured data from user's financial data and storing the collected unstructured data in association with each user; a preprocessing AI unit performing a preprocessing process which classifies the unstructured data on the basis of deep learning, extracts financial data from the classified unstructured data, and structures the financial data; and a marketing AI unit learning the financial data for each user on the basis of preprocessed data preprocessed by the preprocessing AI unit and setting customized financial recommendation information for each user. Accordingly, the apparatus provides an effect of providing customized financial services for each user by collecting various unstructured data generated by financial companies and performing learning on the collected data.

Description

Apparatus and method for recommending artificial intelligence-based financial products using unstructured data {APPARATUS AND METHO FOR RECOMMENDATION OF FINANCIAL PRODUCTS BASED ON ARITIFICAIL INTELLIGENCE USING UNSTRUCRED DATA}

The present invention relates to an artificial intelligence-based financial product recommendation apparatus using unstructured data and a method for providing the same.

As financial products with various and complex structures appear one after another, various technologies are being developed to recommend financial products by judging investment propensity for financial products by using information such as customer's financial product transaction history.

Due to technical limitations, the existing product recommendation model used only standardized data such as customer profiles (gender, age, income level, etc.) and financial transaction details (product subscription history, etc.).

On the other hand, as financial companies secure various computerized systems, unstructured data such as voice, text, and images are increasing more and more.

As a prior art that utilizes such unstructured data, there is Korean Patent Application Laid-Open No. 10-2018-0075852 and the like. However, this prior art has a problem in that it is impossible to utilize various unstructured data generated in actual financial business because text data for arbitrary products and services is only crawled and utilized as unstructured data online.

Korean Patent Publication No. 10-2018-0075852

One technical aspect of the present invention is to solve the problems of the prior art, and it is possible to collect various unstructured data generated in a financial company and learn about it to provide unstructured data that can provide customized financial services for each user. It is to provide an artificial intelligence-based financial product recommendation device and a method for providing the same.

In addition, one technical aspect of the present invention is unstructured data that can improve the performance of an artificial intelligence model by providing automation of the marketing process by using artificial intelligence and re-learning the marketing results through artificial intelligence. To provide an artificial intelligence-based financial product recommendation device and a method for providing the same using

In addition, one technical aspect of the present invention is that by integrating and analyzing face-to-face, non-face-to-face, and outbound marketing, it is possible to switch from the existing product-oriented marketing to customer-centered marketing, and through this, it is possible to utilize unstructured data that enables personalized and customized recommendations. It is to provide an artificial intelligence-based financial product recommendation device and a method for providing the same.

The above objects and various advantages of the present invention will become more apparent from preferred embodiments of the present invention by those skilled in the art.

One technical aspect of the present invention proposes an artificial intelligence-based financial product recommendation apparatus using unstructured data. The artificial intelligence-based financial product recommendation device using the unstructured data collects unstructured data among users' financial data, and an unstructured data collection unit that stores the collected unstructured data in association with each user, deep learning-based unstructured data A preprocessing AI that performs preprocessing to classify, extract finance-related data from the classified unstructured data and standardize it, and financial data for each user based on the preprocessed data preprocessed by the preprocessing AI, and give each user a user It may include marketing artificial intelligence to set customized financial recommendations.

In an embodiment, the user-customized financial recommendation information is selected using a user's financial history and preference information analyzed based on the pre-processing data, and may include at least one of financial products and financial services.

In an embodiment, the artificial intelligence-based financial product recommendation apparatus may further include a customer response collection unit that collects individual user feedback on financial recommendation information provided to each user through a marketing channel.

In an embodiment, the marketing artificial intelligence may set a preferred marketing channel for each user by analyzing a preference for a marketing channel for each user based on the user feedback.

In an embodiment, the pre-processing AI is a pre-learning model that performs learning in advance to learn a language before performing pre-processing, and formalizes unstructured data based on the learned language, data structured by the pre-learning model It may include a classification model for classifying using a previously learned classification criterion and a text summary model that interprets the data standardized by the pre-learning model, and generates main contents in the data as summary data.

In an embodiment, the artificial intelligence-based financial product recommendation device further comprises an unstructured database for classifying and storing the unstructured data collected by the unstructured data collection unit according to the collection means, wherein the unstructured data is Voice of C (VoC). Customer) consultation data, passbook record data, sales memo data, and real-time data can be divided.

In one embodiment, the marketing artificial intelligence analyzes a series of financial events by reading the pre-processing data for each user at daily unit intervals with respect to the pre-processing data pre-processed by the pre-processing artificial intelligence, and analyzing a series of financial events today or in the future. A daily recommendation model that sets recommendation information for each user targeting a specific day, and the preprocessing data preprocessed by the preprocessing artificial intelligence, read the preprocessing data for each user at monthly intervals to obtain a series of finance It may include a monthly recommendation model that analyzes an event and sets recommendation information for each user for this month or a specific month in the future.

In an embodiment, the marketing artificial intelligence may further include a marketing channel model for setting a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on user feedback on the marketing channel. can

In an embodiment, the artificial intelligence-based financial product recommendation device further comprises a structured data collection unit for collecting structured data among users' financial data and storing the collected structured data in association with each user, the pre-processing artificial intelligence can classify structured data and unstructured data using pre-learned classification criteria, and interpret structured data and unstructured data to generate the main contents of the data as summary data.

Another technical aspect of the present invention proposes an AI-based financial product recommendation method using unstructured data. The artificial intelligence-based financial product recommendation method using the unstructured data includes collecting unstructured data among users' financial data, storing the collected unstructured data in association with each user, classifying the unstructured data based on deep learning, Performing pre-processing of extracting financial-related data from the classified unstructured data and standardizing it, and learning financial data for each user based on the pre-processed data pre-processed by the pre-processing artificial intelligence, and providing user-customized financial recommendation information to each user It may include steps to set up.

In an embodiment, the user-customized financial recommendation information is selected using a user's financial history and preference information analyzed based on the pre-processing data, and may include at least one of financial products and financial services.

In one embodiment, the AI-based financial product recommendation method includes collecting individual user feedback on financial recommendation information provided to each user through a marketing channel, and targeting each user based on the user feedback through a marketing channel It may further include the step of setting a preferred marketing channel for each user by analyzing the preference for the.

In an embodiment, the performing of the pre-processing of the formalization includes: learning a language by performing learning in advance before performing the pre-processing, and formalizing the unstructured data based on the learned language; It may include the steps of classifying the data using the pre-learned classification criteria, interpreting the data standardized by the pre-learning model, and generating main contents of the data as summary data.

In one embodiment, the step of setting user-customized financial recommendation information for each user includes reading the pre-processing data for each user at daily intervals for the pre-processing data and analyzing a series of financial events, today or in the future setting recommendation information for each user targeting a specific day of It may include analyzing and setting recommendation information for each user targeting a specific month in this month or in the future.

In an embodiment, the setting of user-customized financial recommendation information for each user may include analyzing preferences for a marketing channel for each individual user based on user feedback on the marketing channel to select a preferred marketing channel for each user. It may further include a step of setting.

The means for solving the above-described problems do not enumerate all the features of the present invention. Various means for solving the problems of the present invention may be understood in more detail with reference to specific embodiments in the following detailed description.

According to one embodiment of the present invention, it is possible to provide customized financial services for each user by collecting various unstructured data generated by a financial company and learning about it.

In addition, according to one embodiment of the present invention, the effect of improving the performance of the artificial intelligence model by providing automation of the marketing process by using artificial intelligence and re-learning the marketing results through artificial intelligence there is

In addition, according to an embodiment of the present invention, by integrating and analyzing face-to-face, non-face-to-face, and outbound marketing, it is possible to switch from the existing product-oriented marketing to customer-centered marketing, and through this, there is an effect that personalized and customized recommendation is possible. .

1 is a view for explaining an application example of an artificial intelligence-based financial product recommendation apparatus using unstructured data according to an embodiment of the present invention.
2 is a diagram for explaining an exemplary computing operating environment of the apparatus for recommending financial products according to an embodiment of the present invention.
3 is a block diagram illustrating an artificial intelligence-based financial product recommendation apparatus according to an embodiment of the present invention.
4 is a block diagram illustrating an embodiment of the pre-processing AI shown in FIG. 3 .
FIG. 5 is a diagram for explaining the unstructured database and preprocessing AI shown in FIG. 3 .
6 is a block diagram illustrating an embodiment of the artificial intelligence marketing shown in FIG. 3 .
FIG. 7 is a diagram for explaining a one-unit recommendation model of artificial intelligence for marketing shown in FIG. 3 .
8 is a block diagram illustrating an artificial intelligence-based financial product recommendation apparatus according to another embodiment of the present invention.
9 is a diagram illustrating an example of a user-customized financial product recommendation screen provided to a user terminal.
10 is a flowchart illustrating an AI-based financial product recommendation method according to an embodiment of the present invention.
11 is a flowchart illustrating an embodiment of step S1020 shown in FIG. 10 .
12 is a flowchart illustrating an embodiment of step S1030 shown in FIG. 10 .

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

However, the embodiment of the present invention may be modified in various other forms, and the scope of the present invention is not limited to the embodiments described below. In addition, the embodiments of the present invention are provided in order to more completely explain the present invention to those of ordinary skill in the art.

That is, the above-described objects, features and advantages will be described later in detail with reference to the accompanying drawings, and accordingly, a person of ordinary skill in the art to which the present invention pertains will be able to easily implement the technical idea of the present invention. In describing the present invention, if it is determined that a detailed description of a known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to indicate the same or similar components.

Also, as used herein, the singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “consisting of” or “comprising” should not be construed as necessarily including all of the various components or various steps described in the specification, some of which components or some steps are It should be construed that it may not include, or may further include additional components or steps.

In addition, various components and sub-components thereof are described below in order to describe an artificial intelligence-based financial product recommendation apparatus using unstructured data according to the present invention. These components and sub-components thereof may be implemented in various forms, such as hardware, software, or a combination thereof. For example, each element may be implemented as an electronic configuration for performing a corresponding function, or may be software itself operable in an electronic system or implemented as a functional element of such software. Alternatively, it may be implemented with an electronic configuration and corresponding driving software.

Although various flowcharts are disclosed to describe the embodiments of the present invention, this is for convenience of description of each step, and each step is not necessarily performed according to the order of the flowchart. That is, each step in the flowchart may be performed simultaneously with each other, performed in an order according to the flowchart, or may be performed in an order opposite to the order in the flowchart.

Hereinafter, an artificial intelligence-based financial product recommendation apparatus using unstructured data and a method for providing the same according to some embodiments of the present invention will be described.

1 is a view for explaining an application example of an artificial intelligence-based financial product recommendation apparatus using unstructured data according to an embodiment of the present invention.

Referring to FIG. 1 , an artificial intelligence-based financial product recommendation device using unstructured data (hereinafter referred to as a 'financial product recommendation device') 100 collects customer data related to a user's financial activity, and collects customer data A financial product corresponding to the user's financial taste can be selected and recommended to the user by applying artificial intelligence to the user.

The financial product recommendation apparatus 100 may use unstructured data as well as structured data as customer data. That is, the financial product recommendation apparatus 100 may perform a preprocessing process of extracting meaningful data from the unstructured data by applying the first artificial intelligence model. Thereafter, the financial product recommendation apparatus 100 may provide user-customized marketing by reflecting the user's financial taste using the second artificial intelligence model.

The customer data of the user may be acquired by various means, and for example, unstructured data such as sales memos or briefs left by the counselor 11 who consulted with the user 1 at the financial window may also be customer data. Alternatively, data of a voice call performed by the user 1 with the customer center may also be customer data as unstructured data. Alternatively, both structured data and unstructured data generated by various activities performed by the user 1 in the financial server through the user terminal 13 may be customer data.

As such, the financial product recommendation apparatus 100 may closely analyze the user's financial taste by deep learning by acquiring customer data of the user, that is, unstructured data and structured data generated in various environments.

Hereinafter, various embodiments of the financial product recommendation apparatus 100 will be described in more detail with reference to FIGS. 2 to 7 .

2 is a diagram for explaining an exemplary computing operating environment of the apparatus for recommending financial products according to an embodiment of the present invention.

2 is intended to provide a general and simplified description of a suitable computing environment in which embodiments may be implemented. Referring to FIG. 2 , as an example of the apparatus 100 for recommending a financial product, a computing device 200 that is a block diagram of an exemplary computing operating environment for an application is illustrated.

Computing device 200 may be any computing device running an application for providing a gallery of pre-built controls for logic scenarios in a database program according to embodiments, and includes at least a processing unit 203 and a system memory ( 201) may be included. A computing device may include a plurality of processing units that cooperate in executing a program. Depending on the exact configuration and type of computing device, system memory 201 may be volatile (eg, RAM), non-volatile (eg, ROM, flash memory, etc.), or a combination thereof. The system memory 201 includes a suitable operating system 202 for controlling the operation of the platform, such as the WINDOWS operating system from Microsoft Corporation. System memory 201 may include one or more software applications, such as program modules, applications, and the like.

The artificial intelligence (AI) model operates based on the operating system and performs learning based on deep learning on big data. Although the artificial intelligence (AI) model is exemplified as a plurality of operations on one computing device, the present invention is not limited thereto.

The computing device may have additional features or functionality. For example, the computing device may include additional data storage devices 204 such as magnetic disks, optical disks, or tape. Such additional storage may be removable storage and/or fixed storage. Computer readable storage media may include volatile and nonvolatile, removable and non-removable media implemented in any method or technology for stored information such as computer readable instructions, data structures, program modules, or other data. The system memory 201 and the storage 204 are only examples of computer-readable storage media. A computer readable storage medium may include RAM (RAM), ROM (ROM), EEPROM, flash memory or other memory technology, CD-ROM, DVD or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage device, or any desired It may include, but is not limited to, any other medium that stores information and that can be accessed by computing device 200 .

input devices 205 of computing devices, such as keyboards, mice, pens, voice input devices, touch input devices, and comparable input devices. Output devices 206 may include, for example, displays, speakers, printers, and other types of output devices. Since these devices are widely known in the art, detailed descriptions thereof will be omitted.

A computing device may include a communication device 207 that allows the device to communicate with other devices via networks 20, such as wired and wireless networks, satellite links, cellular links, local area networks, and comparable mechanisms, such as in a distributed computing environment, for example. may include Communication device 207 is one example of a communication medium, which may include computer readable instructions, data structures, program modules, or other data therein. Illustratively, communication media includes, but is not limited to, wired media such as a wired network or direct connection connection, and wireless media such as acoustic, RF, infrared, and other wireless media.

In FIG. 2 , an example in which the financial product recommendation apparatus 100 is implemented as one computing device 200 has been described, but the present invention is not limited thereto. That is, the financial product recommendation device 100 may be implemented as a plurality of computing devices 200 , or may be implemented as individual computing devices 200 for each function of the financial product recommendation device 100 . etc., can be implemented in various ways.

3 is a block diagram illustrating an artificial intelligence-based financial product recommendation apparatus according to an embodiment of the present invention.

Referring to FIG. 3 , the financial product recommendation apparatus 100 includes an unstructured data collection unit 110 , an unstructured database 120 , a pre-processing artificial intelligence 130 , a pre-processing database 140 , and a marketing artificial intelligence 150 . can According to an embodiment, the financial product recommendation apparatus 100 may further include a customer response collecting unit 160 .

The unstructured data collection unit 110 may collect unstructured data among the user's financial data. For example, the unstructured data collection unit 110 may collect unstructured data of each user by interworking with a server of a financial service company.

The unstructured data may include internal unstructured data existing within a financial service company having the financial product recommendation device 100 and external unstructured data obtained in conjunction with other financial service companies.

The unstructured data includes data with various unstructured characteristics, such as customer service voice files recorded at customer centers, digital log files according to the use of Internet banking or smart banking, sales memos generated at branches, and bankbook summary.

The unstructured data collection unit 110 may store the collected unstructured data in the unstructured database 120 in association with each user.

The pre-processing AI 130 performs pre-processing on the unstructured data stored in the unstructured database 120 . The pre-processing artificial intelligence 130 may perform pre-processing of classifying unstructured data based on deep learning, extracting valid financial-related data from the classified unstructured data, and standardizing it. The pre-processing data generated by the pre-processing artificial intelligence 130 is utilized as structured data, and is stored in the pre-processing database 140 .

The marketing artificial intelligence 150 may learn financial data for each user based on the pre-processing data, and may set user-customized financial recommendation information for each user. The user-customized financial recommendation information is selected using the user's financial history and preference information analyzed based on the pre-processing data, and may include at least one of financial products and financial services.

The marketing artificial intelligence 150 may set user-customized financial recommendation information for each user by performing learning using the pre-processing data for each user.

In an embodiment, the artificial intelligence marketing 150 may determine the user's customized financial recommendation information by analyzing the user's financial data in units of a predetermined period and analyzing the user's financial needs at the current time.

The customer response collection unit 160 may collect individual user feedback on financial recommendation information provided to each user through a marketing channel.

The customer response collection unit 160 may provide the collected user feedback to the marketing AI 150 , and the marketing AI 150 may set a preferred marketing channel for an individual user based on the user feedback.

The marketing artificial intelligence 150 may set a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on the user feedback provided by the customer response collecting unit 160 .

In an embodiment, the artificial intelligence marketing 150 may set a preferred marketing channel for each user by analyzing a preference for a marketing channel in units of a predetermined period. That is, the marketing artificial intelligence 150 analyzes the user's financial data in units of a predetermined period, analyzes the user's financial needs at the current time, determines the user-customized financial recommendation information, and provides the user-customized financial recommendation information. You can set your preferred marketing channels to provide.

Hereinafter, each component of the financial product recommendation apparatus 100 will be described in more detail with reference to FIGS. 4 to 9 .

4 is a block diagram illustrating an embodiment of the preprocessing AI shown in FIG. 3 , and FIG. 5 is a diagram illustrating the unstructured database and the preprocessing AI shown in FIG. 3 .

Referring to FIG. 4 , the pre-processing artificial intelligence 130 includes a pre-learning model 131 , a classification model 132 , a text summary model 132 , and a machine reading comprehension model 134 .

The pre-learning model 131 is a model for pre-learning prior to pre-processing, and is a model for learning and training a main language mainly used in a user environment. For example, when an artificial intelligence-based financial product recommendation device is operated in Korea, the main language is Korean.

The pre-learning model 131 may formalize the unstructured data based on the learned language, ie, convert it into text. For example, the pre-learning model 131 may convert voice data into text based on the learned language.

The classification model 132 may classify the data standardized by the pre-learning model 131 using a pre-learned classification criterion.

5 , the unstructured database 120 may classify and store the unstructured data collected by the unstructured data collection unit 110 according to the collection means. In the illustrated example, the unstructured data is divided into Voice of Customer (VoC) consultation data, passbook record data, sales memo data, and real-time data.

After each of these unstructured data is formalized by the pre-learning model 131 performing learning, the classification model 132 may classify the data structured by the pre-learning model 131 according to a preset criterion. For example, the classification model 132 may classify Voice of Customer (VoC) counseling data into medium classification and emotional classification.

The classification model 132 may interpret the context of the counseling data to determine which classification the counseling data belongs to.

The medium classification is to classify which issue the consultation is about based on the customer consultation content. For example, the classification model 132 can classify consultations on products, service channels, and accident security into medium classification. .

In the emotional classification, customer consultation contents are related to the customer's emotions. For example, the classification model 132 may classify the consultation related to customer emotions, such as subscription, departure, and dissatisfaction, into emotional classification.

In addition, the classification model 132 may classify passbook records into brief classifications and business memos into consultation classifications.

The text summary model 133 may interpret the data standardized by the pre-learning model 131 and generate main contents of the data as summary data.

For example, as shown in FIG. 5 , the text summary model 133 may summarize a core part of consultation content among Voice of Customer (VoC) consultation data.

The machine reading comprehension model 134 may read the data formatted by the pre-learning model 131 . do

For example, the machine reading comprehension model 134 as shown in 5 can read brief classifications, select and analyze the customer's deposit and withdrawal contents among them. Alternatively, the machine reading comprehension model 134 may read the sales memo and analyze the consultation contents of the customer who visited the corresponding branch.

6 is a block diagram illustrating an embodiment of the artificial intelligence marketing shown in FIG. 3 .

Referring to FIG. 6 , the marketing artificial intelligence 150 may include a daily recommendation model 151 , a monthly recommendation model 153 , and a marketing channel model 155 .

The daily recommendation model 151 analyzes a series of financial events by reading the preprocessing data for each user at daily unit intervals, targeting the preprocessed data preprocessed by the preprocessing artificial intelligence 130, and You can set recommended information for each user who is a target for work.

7 is a view for explaining a one-unit recommendation model of the artificial intelligence marketing shown in FIG. 3 , and the daily recommendation model 151 will be described with further reference to FIG. 7 .

As shown in Fig. 7, the daily recommendation model 151 reads the pre-processing data at daily intervals, and receives the pre-processing data from July 3 to July 6 for user A, respectively, the monthly salary of July 3 It can be organized into a series of financial events: deposit, fund search through internet banking on July 4, fund consultation at branch on July 5, and fund consultation through customer center on July 6.

The daily recommendation model 151 may determine that user A has a high degree of interest in fund subscription based on the analyzed series of financial events, and the daily recommendation model 151 recommends fund subscription for user A. It is possible to predict and recommend a fund suitable for user A by analyzing user A's propensity, balance, and already subscribed financial products.

Also, the daily recommendation model 151 may preliminarily recommend a product similar to a fund to user A. That is, the daily recommendation model 151 may select a savings account, which is an accumulation-type product similar to a fund, and analyze user A's propensity, balance, and previously subscribed financial products to recommend a savings savings suitable for user A.

The monthly recommendation model 153 analyzes a series of financial events by reading the pre-processing data for each user at monthly intervals, targeting the pre-processed data pre-processed by the pre-processing artificial intelligence 130 , and this month or future specific You can set recommended information for each user for the month. Since the monthly recommendation model 153 has a similar configuration to the daily recommendation model 151 , a detailed description thereof will be omitted.

The marketing channel model 155 may set a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on user feedback on the marketing channel.

In an embodiment, the marketing channel model 155 may set a preferred marketing channel for each user by analyzing a preference for a marketing channel in units of a predetermined period. That is, the marketing channel model 155 may set a preferred marketing channel for the user on a daily or monthly basis in order to operate in conjunction with the daily recommendation model 151 or the monthly recommendation model 153 .

In the above, an artificial intelligence-based financial product recommendation device that operates on unstructured data has been described. On the other hand, the artificial intelligence-based financial product recommendation apparatus can collect and utilize not only unstructured data but also structured data, and an embodiment in which both unstructured data and structured data are utilized will be described with reference to FIG. 8 .

8 is a block diagram illustrating an artificial intelligence-based financial product recommendation apparatus 101 according to another embodiment of the present invention.

Referring to FIG. 8 , the financial product recommendation device 101 includes an unstructured data collection unit 110 , a collection database 121 , a pre-processing artificial intelligence 130 , a pre-processing database 140 , a marketing artificial intelligence 150 , and a customer. It may include a reaction collection unit 160 and a structured data collection unit 170 .

Here, the content of the unstructured data collection unit 110 to the customer response collection unit 160 is replaced with the description described above with reference to FIGS. 3 to 7 .

The structured data collection unit 170 may collect structured data among the user's financial data. For example, the structured data collection unit 110 may collect unstructured data of each user by interworking with a server of a financial service company.

The structured data collection unit 170 may store the collected unstructured data in the collection database 120 in association with each user.

The collection database 120 separates the structured data and the unstructured data and stores them, and the pre-processing artificial intelligence 130 may perform pre-processing on the structured data and the unstructured data stored in the collection database 120 .

For example, the pre-processing AI 130 classifies structured data and unstructured data using pre-learned classification criteria, and interprets the structured data and unstructured data to generate the main contents of the data as summary data. .

9 is a diagram illustrating an example of a user-customized financial product recommendation screen provided to a user terminal.

The marketing channel shown in FIG. 9 is a push messaging channel targeted to a user terminal. When the user confirms the push notification message, it is possible to easily acquire information on a customized product (recommended product) set for the user.

In the above, an artificial intelligence-based financial product recommendation apparatus according to an embodiment of the present invention has been described with reference to FIGS. 1 to 9 . Hereinafter, an artificial intelligence-based financial product recommendation method according to an embodiment of the present invention will be described with reference to FIGS. 10 to 12 .

Since the AI-based financial product recommendation method to be described below is performed based on the AI-based financial product recommendation apparatus 100, it can be easily understood from the above description based on FIGS. 1 to 9 .

10 is a flowchart illustrating an AI-based financial product recommendation method according to another embodiment of the present invention.

Referring to FIG. 10 , the financial product recommendation apparatus 100 may collect unstructured data among users' financial data and store the collected unstructured data in association with each user ( S1010 ).

The financial product recommendation apparatus 100 may perform pre-processing of classifying unstructured data based on deep learning, extracting finance-related data from the classified unstructured data, and standardizing it (S1020).

The financial product recommendation apparatus 100 may learn financial data for each user based on the pre-processed pre-processed data, and may set user-customized financial recommendation information for each user ( S1030 ).

The user-customized financial recommendation information is selected using the user's financial history and preference information analyzed based on the pre-processing data, and may include at least one of financial products and financial services.

In one embodiment, the apparatus 100 for recommending financial products collects individual users' feedback on financial recommendation information provided to each user through a marketing channel, and based on the user feedback, each user has a preference for a marketing channel. can be analyzed to set preferred marketing channels for each user.

11 is a flowchart illustrating an embodiment of step S1020 shown in FIG. 10 .

Referring to FIG. 11 , the apparatus 100 for recommending a financial product may learn a language by pre-learning before performing pre-processing, and formalize unstructured data based on the learned language ( S1110 ).

The financial product recommendation apparatus 100 may classify the data that has been formalized by the pre-learning model using the pre-learned classification criteria (S1120).

The financial product recommendation apparatus 100 may interpret the data standardized by the pre-learning model and generate main contents of the data as summary data (S1130).

12 is a flowchart illustrating an embodiment of step S1030 shown in FIG. 10 .

Referring to FIG. 12 , the financial product recommendation apparatus 100 analyzes a series of financial events by reading the pre-processing data for each user at daily intervals with respect to the pre-processing data, and targeting a specific day today or in the future. Recommendation information for each user may be set (S1210).

The financial product recommendation device 100 analyzes a series of financial events by reading the pre-processing data for each user at monthly intervals, targeting the pre-processed data pre-processed by the pre-processing artificial intelligence, and this month or a specific month in the future It is possible to set recommended information for each user who is a target (S1220).

In an embodiment, the financial product recommendation apparatus 100 may set a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on user feedback on the marketing channel.

The present invention described above is not limited by the above-described embodiments and the accompanying drawings, but is limited by the claims described below, and the configuration of the present invention may vary within the scope without departing from the technical spirit of the present invention. Those of ordinary skill in the art to which the present invention pertains can easily recognize that it can be changed and modified.

10 : user
11: Counselor
12, 13: user terminal
100: Artificial intelligence-based financial product recommendation device
110: unstructured data collection unit 120: unstructured database
130: preprocessing artificial intelligence 140: preprocessing database
150: marketing artificial intelligence 160: customer response collection unit
200: computing device
201: system memory 202: operating system
203 processing unit 204 storage
205: input device 206: output device
207: communication device

Claims (15)

an unstructured data collection unit that collects unstructured data among users' financial data, and stores the collected unstructured data in association with each user;
Pre-processing artificial intelligence that classifies unstructured data based on deep learning and performs pre-processing of extracting financial-related data from the classified unstructured data and making it standardized; and
Marketing artificial intelligence for learning financial data for each user based on the pre-processing data pre-processed by the pre-processing artificial intelligence and setting user-customized financial recommendation information for each user;
Artificial intelligence-based financial product recommendation device using unstructured data including
The method of claim 1, wherein the user-customized financial recommendation information
At least one of financial products and financial services to be selected using the user's financial history and preference information analyzed based on the pre-processing data
Artificial intelligence-based financial product recommendation device using unstructured data including
According to claim 1, wherein the artificial intelligence-based financial product recommendation device,
a customer response collecting unit that collects individual user feedback on financial recommendation information provided to each user through a marketing channel;
Artificial intelligence-based financial product recommendation device using unstructured data that further includes
The method of claim 3, wherein the marketing artificial intelligence
Setting a preferred marketing channel for each user by analyzing the preference for the marketing channel for each user based on the user feedback
Artificial intelligence-based financial product recommendation device using unstructured data.
The method of claim 1, wherein the pre-processing artificial intelligence
a pre-learning model that learns a language by performing pre-processing in advance, and formalizes unstructured data based on the learned language;
a classification model for classifying the data standardized by the pre-learning model using pre-learned classification criteria; and
a text summary model that interprets data formatted by the pre-learning model and generates main contents of the data as summary data;
Artificial intelligence-based financial product recommendation device using unstructured data including
According to claim 1, wherein the artificial intelligence-based financial product recommendation device,
an unstructured database for classifying and storing the unstructured data collected by the unstructured data collection unit according to the collection means;
further comprising,
The unstructured data is
VoC (Voice of Customer) consultation data, passbook record data, sales memo data, and real-time data
Artificial intelligence-based financial product recommendation device using unstructured data.
According to claim 1, wherein the marketing artificial intelligence
Based on the pre-processed data pre-processed by the pre-processing artificial intelligence, the pre-processing data for each user is read at daily intervals to analyze a series of financial events, and recommendations for each user targeting a specific day today or in the future Daily recommendation model to set information; and
Based on the pre-processed data pre-processed by the pre-processing artificial intelligence, the pre-processing data for each user is read at monthly intervals to analyze a series of financial events, and recommendations for each user for this month or a specific month in the future Monthly recommendation model to set information;
Artificial intelligence-based financial product recommendation device using unstructured data including
The method of claim 7, wherein the marketing artificial intelligence
a marketing channel model for setting a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on user feedback on the marketing channel;
Artificial intelligence-based financial product recommendation device using unstructured data that further includes
According to claim 1, wherein the artificial intelligence-based financial product recommendation device,
a structured data collecting unit that collects structured data among users' financial data, and stores the collected structured data in association with each user;
further comprising,
The preprocessing artificial intelligence
It classifies structured data and unstructured data using pre-learned classification criteria, and interprets structured data and unstructured data to generate the main contents of the data as summary data.
Artificial intelligence-based financial product recommendation device using unstructured data.
collecting unstructured data among the user's financial data, and storing the collected unstructured data in association with each user;
Classifying the unstructured data based on deep learning, extracting financial-related data from the classified unstructured data, and performing pre-processing for standardization; and
learning financial data for each user based on the pre-processed pre-processed data, and setting user-customized financial recommendation information for each user;
Artificial intelligence-based financial product recommendation device using unstructured data including
The method of claim 10, wherein the user-customized financial recommendation information
An artificial intelligence-based financial product recommendation method using unstructured data that is selected using the user's financial history and preference information analyzed based on the pre-processing data, and includes at least one of financial products and financial services.
The method of claim 10, wherein the AI-based financial product recommendation method comprises:
collecting individual users' feedback on financial recommendation information provided to each user through a marketing channel; and
setting a preferred marketing channel for each user by analyzing a preference for a marketing channel for each user based on the user feedback;
An artificial intelligence-based financial product recommendation method using unstructured data that further includes
11. The method of claim 10, wherein the performing of the pre-processing of the shaping comprises:
learning a language by pre-learning before performing pre-processing, and formalizing unstructured data based on the learned language;
Classifying the data formalized by the pre-learning model using the pre-learned classification criteria; and
interpreting the data standardized by the pre-learning model, and generating main contents of the data as summary data;
AI-based financial product recommendation method using unstructured data including
The method of claim 10, wherein the setting of user-customized financial recommendation information for each user comprises:
Analyzing a series of financial events by reading the pre-processing data for each user at daily intervals with respect to the pre-processing data, and setting recommendation information for each user targeting a specific day today or in the future; and
Analyzing a series of financial events by reading the pre-processing data for each user at monthly intervals with respect to the pre-processing data, and setting recommendation information for each user for this month or a specific month in the future;
AI-based financial product recommendation method using unstructured data including
The method of claim 14, wherein the setting of user-customized financial recommendation information for each user comprises:
setting a preferred marketing channel for each user by analyzing a preference for a marketing channel for each individual user based on user feedback on the marketing channel;
An artificial intelligence-based financial product recommendation method using unstructured data that further includes

Images