CN113849731A

CN113849731A - Information pushing method, device, equipment and medium based on natural language processing

Info

Publication number: CN113849731A
Application number: CN202111105834.5A
Authority: CN
Inventors: 曹雷; 徐集优; 常孟可; 史凯旭
Original assignee: Ping An Bank Co Ltd
Current assignee: Ping An Bank Co Ltd
Priority date: 2021-09-22
Filing date: 2021-09-22
Publication date: 2021-12-28
Anticipated expiration: 2041-09-22
Also published as: CN113849731B

Abstract

The application relates to an artificial intelligence technology, and provides an information pushing method, an information pushing device, information pushing equipment and an information pushing medium based on natural language processing, wherein the method comprises the following steps: acquiring a historical record to be analyzed on a target application program of a target client and at least two analyzed historical sub-evaluation values; analyzing the historical records based on a natural language processing model to obtain at least two reference sub-evaluation values; updating the historical sub-evaluation value based on the reference sub-evaluation value to obtain a target sub-evaluation value; weighting the target sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value; selecting a target service from a resource pool of a target application program based on the target sub-evaluation value and the target comprehensive evaluation value; and pushing the pushing information of the target service to the target client. By the method and the device, information pushing can be performed based on the evaluation information of the client, and the success rate of service pushing can be improved.

Description

Information pushing method, device, equipment and medium based on natural language processing

Technical Field

The application relates to the technical field of artificial intelligence, and mainly relates to an information pushing method, device, equipment and medium based on natural language processing.

Background

With the development of computers and the internet, many services can be performed on the internet. Such as trading transactions of stocks, funds, futures, bonds, derivatives, brokerages, transfers, etc. At present, the activity page in the application program corresponding to the transaction service is generally uniform content, which results in low success rate of service push.

Disclosure of Invention

The embodiment of the application provides an information pushing method, an information pushing device, information pushing equipment and an information pushing medium based on natural language processing, information pushing can be carried out based on evaluation information of a client, and the success rate of service pushing is improved.

In a first aspect, an embodiment of the present application provides an information pushing method based on natural language processing, where:

acquiring a historical record to be analyzed on a target application program of a target client and at least two analyzed historical sub-evaluation values;

analyzing the historical records based on a natural language processing model to obtain at least two reference sub-evaluation values;

updating the historical sub-evaluation value based on the reference sub-evaluation value to obtain a target sub-evaluation value;

weighting the target sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value;

selecting a target service from a resource pool of the target application program based on the target sub-evaluation value and the target comprehensive evaluation value;

and pushing the pushing information of the target service to the target client.

In a second aspect, an embodiment of the present application provides an information pushing apparatus based on natural language processing, where:

the system comprises an acquisition unit, a storage unit and a processing unit, wherein the acquisition unit is used for acquiring a historical record to be analyzed on a target application program of a target client and at least two analyzed historical sub-evaluation values;

the analysis unit is used for analyzing the historical record based on a natural language processing model to obtain at least two reference sub-evaluation values;

an updating unit, configured to update the historical sub-evaluation value based on the reference sub-evaluation value to obtain a target sub-evaluation value;

the weighting unit is used for weighting the target sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value;

a selecting unit, configured to select a target service from a resource pool of the target application based on the target sub-evaluation value and the target comprehensive evaluation value;

and the pushing unit is used for pushing the pushing information of the target service to the target client.

In a third aspect, an embodiment of the present application provides a computer device, including a processor, a memory, a communication interface, and one or more programs, where the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for some or all of the steps described in the first aspect.

In a fourth aspect, the present application provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, where the computer program makes a computer execute to implement part or all of the steps described in the first aspect.

The embodiment of the application has the following beneficial effects:

after the information pushing method, the information pushing device, the information pushing equipment and the information pushing medium based on natural language processing are adopted, the historical records to be analyzed on the target application program of the target client and at least two historical sub-evaluation values obtained through analysis are obtained firstly. And analyzing the historical record based on the natural language processing model to obtain at least two reference sub-evaluation values, and updating the historical sub-evaluation values based on the reference sub-evaluation values to obtain target sub-evaluation values. And weighting the target sub-evaluation value and a preset sub-weight value of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value. And then selecting a target service from a resource pool of the target application program based on the target sub-evaluation value and the target comprehensive evaluation value, and finally pushing the pushing information of the target service to a target client. Therefore, information pushing can be carried out based on the historical record to be analyzed and the at least two historical sub-evaluation values obtained through analysis, and the success rate of service pushing is improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Wherein:

fig. 1 is a schematic flowchart of an information push method based on natural language processing according to an embodiment of the present disclosure;

fig. 2 is a schematic structural diagram of an information pushing apparatus based on natural language processing according to an embodiment of the present application;

fig. 3 is a schematic structural diagram of a computer device according to an embodiment of the present application.

Detailed Description

In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art without any inventive work according to the embodiments of the present application are within the scope of the present application.

The terms "first," "second," and the like in the description and claims of the present application and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.

The network architecture applied by the embodiment of the application comprises a server and electronic equipment. The number of the electronic devices and the number of the servers are not limited in the embodiment of the application, and the servers can provide services for the electronic devices at the same time. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like. The server may alternatively be implemented as a server cluster consisting of a plurality of servers.

The electronic device may be a Personal Computer (PC), a notebook computer, or a smart phone, and may also be an all-in-one machine, a palm computer, a tablet computer (pad), a smart television playing terminal, a vehicle-mounted terminal, or a portable device. The operating system of the PC-side electronic device, such as a kiosk or the like, may include, but is not limited to, operating systems such as Linux system, Unix system, Windows series system (e.g., Windows xp, Windows 7, etc.), Mac OS X system (operating system of apple computer), and the like. The operating system of the electronic device at the mobile end, such as a smart phone, may include, but is not limited to, an operating system such as an android system, an IOS (operating system of an apple mobile phone), a Window system, and the like.

The electronic device may install and run the application program, and the server may be a server corresponding to the application program installed in the electronic device, and provide an application service for the application program. The application program may be a single integrated application software, or an applet embedded in another application, or a system on a web page, etc., which is not limited herein.

The target application involved in the embodiment of the present application may be an application of a bank, or may be an application of an insurance, or may be an application of a security, etc., for providing a transaction service. The business corresponding to the transaction service provided by the target application program can comprise a live business, a regular business, a transfer business, a financing business, an insurance business or an ecological business and the like.

The current business refers to the unmanaged amount of money and generated interest in the bank card bound in the target application program. The amount may be revenue transferred in the case where a preset condition (e.g., expiration, receipt of a transfer instruction, etc.) is satisfied, for example, payroll, other transfer, financing service, insurance service, and periodic service. The regular services may include services corresponding to regular durations of three months, six months, one year, three years, five years, etc., each regular service corresponding to a different amount of interest.

Transfer services typically include roll-out services, i.e., services that transfer money to others. The transfer can be carried out through bank cards of other people, or after the financial service of the third party application opens the authority, the transfer can be carried out through account numbers or bound mobile phone numbers of the third party application, and the like.

The financial service may include a financial service provided by a bank corresponding to the target application program, and may further include a financial service provided by a third party. The financial services may include fund services, stock services, gold services, etc., and may also include insurance services of a financial nature, etc.

The insurance services may include insurance services, e.g., security insurance, etc., for the customer using the target application. And the method can also comprise the insurance business of paying by the client by using the bank card bound in the target application program, and the insured person of the insurance business can be the principal or other people such as parents, spouses, children and the like, and the insurance business is provided by the insurance company. And may also include government provided medical insurance and the like.

The ecological services may include peripheral services introduced into the target application, such as life payment, telephone charge recharge, commemorative coin reservation, active fitness, pension, housing accumulation, preferential activities of people's business or third party applications, etc., which may be applied to many aspects of customer life. It should be noted that the peripheral service may be associated with a location, and the location may be a region corresponding to the card opening, a region selected by the customer, or a region where the customer is currently located.

The embodiment of the application provides an information pushing method based on natural language processing, which can be executed by an information pushing device based on natural language processing. The device can be realized by software and/or hardware, can be generally integrated in electronic equipment or a server, can carry out information push based on the evaluation information of a client, and is favorable for improving the success rate of service push.

Referring to fig. 1, fig. 1 is a schematic flowchart illustrating an information push method based on natural language processing according to the present application. Taking the application of the method to a server as an example for illustration, the method includes the following steps S101 to S106, where:

s101: and acquiring a historical record to be analyzed on the target application program of the target client and at least two analyzed historical sub-evaluation values.

In the embodiment of the application, the target client can be any client registered in the target application program. Or may be a client currently running the target application, or may be a client that has just submitted a service in the target application, etc., which is not limited herein. The target application can refer to the foregoing description, and is not described herein again.

The historical sub-evaluation value may be included in the historical evaluation information obtained by the last data analysis. The historical evaluation information may further include a historical comprehensive evaluation value, which is a numerical value obtained by weighting each historical sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the historical sub-evaluation value. The history sub-evaluation value, the history comprehensive evaluation value, and the preset weight may refer to the description of the target sub-evaluation value and the target comprehensive evaluation value acquisition method described later.

The time interval between the last data analysis and the current data analysis may be a specified time period suitable for all persons, for example, one month, or one week, 6 months, etc. Or may be a specified duration of time suitable for the target customer, and may be determined based on the active value of the target customer using the target application. The active value is used to evaluate the frequency of use of the target application by the target client, and the description of the active value refers to the description of the sub-evaluation value later. It can be understood that data analysis is performed based on the historical records to be analyzed corresponding to the time interval determined by the active value of the target client and the analyzed historical sub-evaluation values, and the effectiveness of pushing the service to the target client can be improved.

The time interval may alternatively be determined based on attribute information of the transaction service that the target customer has committed. The transaction traffic may include payroll income, periodic financing, due traffic, and traffic just submitted, among others. It will be appreciated that payroll time is typically a fixed time of a month, and that periodic financing time is also fixed, for example, a month, or a week, or a day, etc. The time interval for analyzing the service information of the target customer is determined based on the transaction time of the transaction service, so that the periodicity of the push service can be improved. And after payroll income is credited and income of expired business is credited, or after the targeted customer submits a new business, the targeted customer may consider other businesses. In this case, information push is performed, and the success rate of push can be improved. And after the service is submitted, the value of the target client is reevaluated, so that the effectiveness of pushing the service to the target client can be improved.

The historical record to be analyzed may include details of the target customer's transactions on the target application over a time interval so that information about each of the target customer's expenses and incomes may be obtained. The transaction details may be transaction details of a transaction order submitted by the target customer when using the target application, and may also include transaction details of a transaction order generated by making a payment using a bank card or payment account bound in the target application. The target history record may alternatively include a browsing record and/or a searching record of the target client on the target application program in a time interval, and the like, which is not limited herein.

And if the target client is a new user, the historical evaluation information of the target client is null. The time interval may be the length of time from the first time the target client enters the target application to the time the target application exits, and the history may be all records from the first time the target client enters the target application to the time the target application exits.

S102: and analyzing the historical record based on a natural language processing model to obtain at least two reference sub-evaluation values.

In this embodiment of the application, the preset evaluation dimensions corresponding to the history sub-evaluation value and the target sub-evaluation value include a contribution dimension, an active dimension, a loyalty dimension, a qualification dimension, an ecological dimension, and the like, which is not limited herein. And the sub-evaluation value corresponding to the contribution dimension is a contribution value and is used for evaluating profits which can be generated by the target client in the process of using the target application program. And the sub-evaluation value corresponding to the active dimension is an active value and is used for evaluating the frequency of the target client accessing the target application program. And the sub-evaluation value corresponding to the loyalty dimension is loyalty value and is used for evaluating the stability of the target application program used by the target client. And the sub-evaluation value corresponding to the qualification dimension is the qualification value and is used for evaluating the fund distribution of the non-frozen fund of the target customer. And the sub-evaluation value corresponding to the ecological dimension is ecological value and is used for evaluating the purchasing power of the target customer for purchasing the ecological business and the frequency of browsing the ecological business.

The natural language processing model may be based on a natural language processing algorithm, and may use a jieba word segmentation tool, or a word vector model of word2vec, etc. for parsing the text to obtain the part of speech (e.g., two categories of nouns and verbs, as well as names of people, places, names of organizations, etc., or verb by-pass, verb by name, etc.) and word senses corresponding to each word or word in the text. Further, the natural language processing model may also be based on a classification algorithm, such as a decision tree, for analyzing the client value corresponding to the parsed word.

In the embodiment of the present application, the natural language processing model may be configured to analyze the history record to obtain information in the history record, and then analyze the information to obtain at least two reference sub-evaluation values. The method for analyzing the history record is not limited, and the value corresponding to the numerical value of each section of data can be preset. Please refer to table 1 below, which illustrates the cross border settlement in about 6 months.

TABLE 1

As shown in table 1, the value of 18926 corresponds to the target client when the number of cross border settlements is less than or equal to 33 in the last 6 months. The corresponding value of the target client is 38648 when the cross border settlement stroke number is larger than 33 and smaller than or equal to 90 in the last 6 months. The corresponding value of the target client is 77440 when the cross border settlement stroke number is more than 90 in the last 6 months. Therefore, cross-border trading orders of nearly 6 months can be extracted from the history records based on the natural language processing model, and then the number of cross-border settlement strokes of nearly 6 months is counted. And then the value corresponding to the cross-border settlement stroke number of the target client in the last 6 months can be obtained based on the table 1.

In one possible example, step S102 may include the following steps A1-A4, wherein:

a1: and acquiring a data type corresponding to a preset evaluation dimension based on the subimage model of the preset evaluation dimension corresponding to the historical subinterval value.

In the embodiment of the application, the sub-portrait model of the preset evaluation dimension corresponding to the historical sub-evaluation value is used for describing the characteristics of the target client of the preset evaluation dimension corresponding to the historical sub-evaluation value. The sub-portrait model comprises data nodes corresponding to data types required by data analysis, and the data types corresponding to the data nodes in the sub-portrait model can be used as data types corresponding to preset evaluation dimensions.

The construction method of the sub-portrait model with the preset evaluation dimension is not limited, and in one possible example, historical sub-data corresponding to the preset evaluation dimension is obtained from the basic information and the transaction data of the target customer; and connecting data nodes corresponding to the historical subdata based on the connection relation between the historical subdata to obtain the subimage model with the preset evaluation dimension.

In the embodiment of the present application, the basic information of the target client may include the name, age, contact phone, home address, academic calendar, work experience, contact, asset information, etc. of the target client. The transaction data for the target customer may include a transaction order for the target customer in the target application. The connection relationship between the history sub data may be determined based on whether an association relationship exists between the two history sub data and an association value corresponding to the association relationship.

It is understood that, in this example, the history sub-data corresponding to the preset evaluation dimension is acquired from the basic information and the transaction data of the target customer. And connecting data nodes corresponding to the historical subdata based on the connection relation between the historical subdata to obtain a subimage model with preset evaluation dimensions. Therefore, the portrait model is adopted to describe the data of the preset evaluation dimension of the target client, and the accuracy of obtaining the characteristics of the target client is improved. Furthermore, the historical sub-evaluation value may also be acquired by the connection relationship between the respective data nodes.

A2: and acquiring the subdata corresponding to the data type from the historical record based on a natural language processing model.

In the embodiment of the present application, the sub data corresponding to the data type is data for evaluating a reference sub evaluation value corresponding to a preset evaluation dimension. For example, the subdata corresponding to the contribution dimension may include transaction details in the target history record, and the subdata corresponding to the ecological dimension may include transaction details related to the ecological business in the transaction details. The sub-data corresponding to the qualification dimension may include the total assets in the bank card of the target customer and the detailed information of each financing service. The child data corresponding to the active dimension may include browsing records and/or search records of the target client. The sub-data corresponding to the loyalty dimension may include a usage record of the user using the target application. The historical records can be analyzed based on the natural language processing model, and the data types corresponding to the preset evaluation dimensions and the subdata corresponding to the preset evaluation dimensions are obtained.

A3: and calculating the sub data based on the operation expression corresponding to the sub portrait model to obtain the evaluation value corresponding to the sub data.

In the embodiment of the application, the connection relationship between the data nodes is described in the child portrait model, and an operation expression between data of data types corresponding to the data nodes can be obtained based on the connection relationship, so that child data corresponding to a preset evaluation dimension in a history record can be calculated based on the operation expression, and an evaluation value corresponding to the child data is obtained.

A4: and weighting the evaluation value corresponding to the subdata and a preset subduer value to obtain a reference subdocument value corresponding to the preset evaluation dimension.

The preset sub-weight value (or preset sub-weight) corresponding to the sub-data is not limited, and can be a preset numerical value. For example, the preset sub-weight of the cross-border settlement stroke number in the last 6 months is 0.5, the preset sub-weight of the financing stroke number in the last 6 months is 0.5, the preset sub-weight of the transfer stroke number in the last 6 months is 0.1, and the like.

In one possible example, the following steps may be further included: determining an adjustment parameter of the subdata; and adjusting the industry weight of the subdata based on the adjustment parameter to obtain the preset subdata weight.

In the embodiment of the present application, the industry weight of the subdata is a weight usually set for the subdata in the financial industry. The internal weight value can be obtained through the numerical value on the connecting line between the data node corresponding to the subdata in the sub-model and the data node with the connection relation of the data node.

The adjustment parameters are used to adjust the industry weight. The adjustment parameter may be acquired based on a ratio between the data amount of the sub data and the data amount corresponding to the history sub evaluation value, or may be acquired based on a client type of the target client (for example, a client class, a work industry of the client, or the like). The adjustment parameter may be the sum of the industry weights corresponding to all the subdata, a preset subdeigh value, or a ratio between the industry weight and the adjustment parameter. Referring to table 2 below, the sub-data corresponding to the contribution dimension includes the cross-border settlement number of the last 6 months, the financing number of the last 6 months and the transfer number of the last 6 months.

TABLE 2

Type of sub data	Industry weight	Presetting the weight
			Cross-border settlement for nearly 6 months	0.5	0.4545＝0.5/sum(R2)1.1
The financial management strokes in about 6 months	0.5	0.4545＝0.5/sum(R2)1.1
			Transfer strokes in nearly 6 months	0.1	0.0909＝0.1/sum(R2)1.1

Wherein the internal weight of the cross-border settlement strokes in the last 6 months is 0.5, the internal weight of the financing strokes in the last 6 months is 0.5, and the internal weight of the transfer strokes in the last 6 months is 0.1. The adjustment parameter is the sum of the internal weights of the above three subdata, i.e. the calculation formula of the adjustment parameter is 1.1 as 0.5+0.5+ 0.1. The preset sub-weight is shown in table 2, and the preset sub-weight of the cross-border settlement stroke number in the last 6 months is equal to the ratio 0.4545 between the industry weight (0.5) and the adjustment parameter (1.1). If the internal weight value between the amount of financing in the last 6 months and the amount of cross-border settlement in the last 6 months is equal, the preset sub-weight value is also equal and equal to 0.4545. The preset sub-weight of the transfer strokes in the last 6 months is equal to the ratio 0.0909 between the industry weight (0.1) and the adjustment parameter (1.1).

It can be understood that, in this example, the industry weight of the sub-data is adjusted based on the adjustment parameter of the sub-data to obtain the preset sub-weight of the sub-data, which can improve the accuracy of setting the preset sub-weight and is beneficial to improving the accuracy of calculating the reference sub-evaluation value.

It can be understood that after the evaluation value corresponding to each sub-data is obtained, a weighted calculation may be performed based on the preset sub-weight of the sub-data to obtain a reference sub-evaluation value corresponding to the preset evaluation dimension. If the preset evaluation dimension corresponds to n pieces of sub data, the reference sub evaluation value S corresponding to the preset evaluation dimension may be as shown in the following formula (2).

Wherein x is_nAn evaluation value corresponding to the nth sub data, α_nAnd the preset sub-weight value of the nth sub-data is obtained.

It is understood that in steps a 1-a 4, the data type corresponding to the preset evaluation dimension is obtained based on the sprite model of the preset evaluation dimension corresponding to the historical sub-evaluation value. And acquiring subdata corresponding to the data type from the historical record based on the natural language processing model. And then, calculating the sub data based on the operation expression corresponding to the sub image model to obtain the evaluation value corresponding to the sub data. And weighting the evaluation value corresponding to the sub-data and the preset sub-weight value to obtain a reference sub-evaluation value corresponding to the preset evaluation dimension. Therefore, the accuracy and comprehensiveness of obtaining the client characteristics can be improved.

S103: and updating the historical sub-evaluation value based on the reference sub-evaluation value to obtain a target sub-evaluation value.

In the embodiment of the present application, the target sub-evaluation value is an evaluation value obtained by updating the history sub-evaluation value based on the reference sub-evaluation value, so that the real-time performance of the features of the target client can be improved. The present application is not limited to the method for updating the history sub-evaluation value, and in a possible example, the step S103 may include the following steps: calculating a first ratio between the time length corresponding to the historical record and the time length corresponding to the historical sub-evaluation value; calculating a second ratio between the data quantity of the sub data and the data quantity corresponding to the historical sub evaluation value; a target sub-evaluation value is calculated based on the first ratio, the second ratio, the reference sub-evaluation value, and the history sub-evaluation value.

In the embodiment of the present application, the first ratio is a ratio between a time length corresponding to the history and a time length corresponding to the history sub-evaluation value, that is, a ratio between a time length of unanalyzed data and a time length of analyzed data. The second ratio is a ratio between the data amount of the sub data corresponding to the preset evaluation dimension in the history record and the data amount corresponding to the history sub evaluation value, that is, a ratio between the data amount that has not been analyzed and the data amount that has been analyzed.

Optionally, the adjustment parameter is calculated based on the first ratio and the second ratio, a product between the adjustment parameter and the historical sub-evaluation value is obtained, and a sum of the product and the reference sub-evaluation value is obtained as the target evaluation value. In this way, the data volumes corresponding to the history sub-evaluation value and the reference sub-evaluation value can be balanced, which is beneficial to improving the accuracy of calculating the target sub-evaluation value.

It is to be understood that, in this example, the first ratio and the second ratio are calculated from the viewpoint of time and data amount, respectively, and the target sub-evaluation value is calculated based on the first ratio, the second ratio, the history sub-evaluation value, and the reference sub-evaluation value. In this way, the data volumes corresponding to the history sub-evaluation value and the reference sub-evaluation value can be balanced, which is beneficial to improving the accuracy of calculating the target sub-evaluation value.

S104: and weighting the target sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value.

The preset weight corresponding to the preset evaluation dimension may refer to the description of the preset sub-weight of the sub-data, or may be adjusted based on an adjustment parameter of the preset evaluation dimension to obtain the preset weight corresponding to the preset evaluation dimension, which is not described herein again. Therefore, the accuracy of acquiring the target comprehensive evaluation value can be improved by weighting and calculating the target sub-evaluation values.

S105: and selecting target services from the resource pool of the target application program based on the target sub-evaluation value and the target comprehensive evaluation value.

In an embodiment of the present application, the resource pool of the target application includes transaction services supported in the target application. The target service is a service selected based on the target sub-evaluation value and the target comprehensive evaluation value. The method for selecting the target service is not limited in the present application, and in a possible example, the step S105 may include the following steps B1 to B5:

b1: and selecting reference services corresponding to the target sub-evaluation value and the target comprehensive evaluation value from the resource pool of the target application program.

In the embodiment of the present application, the reference service is a service selected based on the target sub-evaluation value and the target comprehensive evaluation value. The method for selecting the reference service is not limited, and the sub-evaluation value required by the service in the resource pool is greater than or equal to the target sub-evaluation value, and the service corresponding to the comprehensive evaluation value required by the service is greater than or equal to the target comprehensive evaluation value can be used as the reference service.

B2: and if the number of the reference services is greater than or equal to 2, determining the recommendation probability of the reference services based on the service attributes of the reference services.

In the embodiment of the present application, the service attribute of the reference service may include a name, a type, revenue information, revenue risk, and the like of the reference service. The type of the service may include a current service, a periodic service, a transfer service, a financing service, an insurance service, an ecological service or a loan service, etc., as described above. The profit information of the service refers to the profit or profit range that the customer can obtain to purchase the service. For example, interest on the live business is 0.3%, interest on the one-year-scheduled business is 1.75%, and interest on the financing business is 4.6%, etc. The profit risk of a business refers to the level of risk that a customer needs to bear to purchase the business, and may be determined based on the type of business, for example, different financial businesses have different profit risks. The risk of revenue may be described in high, medium or low form and is not limited herein.

The recommendation probability is used to describe the recommendation value of the reference service itself. It can be understood that if the number of the reference services is greater than or equal to 2, it indicates that there are multiple services capable of being pushed, so that a target service can be selected from the reference services for pushing, so as to improve the pushing efficiency. The recommendation probability may be determined based on the pushed age and evaluation information of the reference service. The push aging can include the activity remaining time of the reference service, the qualification for purchasing the reference service and the like. The rating information may include review information for the reference service in the target application or on other applications. It can be understood that the recommendation probability of the reference service is determined based on the push timeliness and the evaluation information of the reference service, and the success rate of the push service can be improved.

B3: determining a preference value for the reference service based on the history.

In the embodiment of the present application, the preference value is used to describe the probability that the target customer may purchase the reference service. The determination may be made based on a matching value between the attribute information of the purchased service of the target customer and the attribute information of the reference service. In one possible example, step B3 may include the steps of: selecting a target historical record corresponding to the attribute information of the reference service from the historical records; acquiring the average browsing duration and the purchasing success rate of the target customer based on the target history record; and calculating the preference value of the reference service based on the average browsing duration and the purchase success rate.

The target service information refers to a history record which is the same as the attribute information of the reference service in the history record. The same means that the information such as the type, income information and income risk of the business are equal. It can be understood that the accuracy of obtaining the recommended value can be improved by obtaining the preference value of the reference service based on the target history.

The average browsing duration refers to an average value of browsing durations for the target client to browse the service identical to the attribute information of the reference service. The purchase success rate is a success rate of purchasing a service identical to the attribute information of the reference service, and may be calculated by a ratio between the number of purchases and the total number.

The preference value can be determined based on a numerical value obtained by weighted calculation of the average browsing duration and the purchase success rate; or may be calculated based on a product between the sub-preference value corresponding to the average browsing duration and the purchase success rate, and the like, which is not limited herein.

It is understood that, in this example, the target history record corresponding to the attribute information of the reference service is selected from the history records. And then the preference value of the reference service is calculated based on the average browsing duration and the purchasing success rate of the target customer acquired by the target historical record, so that the accuracy of determining the preference value can be improved.

B4: and calculating the recommended value of the reference service based on the recommended probability and the preference value.

B5: and selecting a target service from the reference services based on the recommended value.

In this embodiment of the application, the recommendation value may be a numerical value obtained by weighting the recommendation probability and the preference value, or may be a product between the recommendation probability and the preference value, and the like, which is not limited herein. The target service may be a reference service whose recommended value is greater than a preset threshold, or the first N reference services whose recommended value is the largest, and the like.

It is understood that, in steps B1-B5, reference services corresponding to the target sub-rating and the target comprehensive rating are first selected from the resource pool of the target application. When the number of reference services is greater than or equal to 2, a recommendation probability of the reference services may be determined based on the service attributes of the reference services, and a preference value of the reference services may be determined based on the history. And then selecting the target service from the reference services based on the recommendation value obtained by calculating the recommendation probability and the preference value. Therefore, the accuracy of selecting the target service can be improved according to the recommendation probability of the reference service and the preference value which can represent the preference of the target client, and the success rate of service pushing is favorably improved.

S106: and pushing the pushing information of the target service to the target client.

In the embodiment of the present application, the push information refers to information that is pushed to a target client by a target service, and may be in the form of characters or images. The push information may include information or a page of the target application for presentation, and the like, which is not limited herein.

In one possible example, step S106 may include the steps of: acquiring keywords and/or pushed images of the target service based on the attribute information of the target service; acquiring push information of the target service based on the keyword and/or the push image; and pushing the push information to the target client.

The keywords are used for describing the salient features of the target service, such as slogans, activity level, income value and the like, and can be obtained based on the attribute information of the target service. The pushed image is used for describing important information of the target service, such as the name, icon, keyword and the like of the target service, and can also be obtained based on attribute information of the target service. The pushed image may be an image used by all people, or may be an image used by the target customer, for example, a color scheme modified based on the target customer's preferences, or a text of interest to the target customer may be highlighted, etc. The push information may include words corresponding to the keywords, or an image obtained by adjusting the push image based on the target client's preferences and keywords, or the like.

It can be understood that, in this example, the pushing information of the target service is pushed to the target client based on the keyword and/or the pushing image of the target service, so that the target client obtains the basic information of the target service through the pushing information, and the pushing effectiveness can be improved. And the target client can quickly know the basic information of the target service through the keywords and/or the push image so as to improve the push efficiency.

In the method shown in fig. 1, a history record to be analyzed and at least two analyzed history sub-evaluation values of a target client on a target application program are obtained. And analyzing the historical record based on the natural language processing model to obtain at least two reference sub-evaluation values, and updating the historical sub-evaluation values based on the reference sub-evaluation values to obtain target sub-evaluation values. And weighting the target sub-evaluation value and a preset sub-weight value of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value. And then selecting a target service from a resource pool of the target application program based on the target sub-evaluation value and the target comprehensive evaluation value, and finally pushing the pushing information of the target service to a target client. Therefore, information pushing can be carried out based on the historical record to be analyzed and the at least two historical sub-evaluation values obtained through analysis, and the success rate of service pushing is improved.

The method of the embodiments of the present application is set forth above in detail and the apparatus of the embodiments of the present application is provided below.

Referring to fig. 2, fig. 2 is a schematic structural diagram of an information pushing apparatus based on natural language processing according to the present application, consistent with the embodiment shown in fig. 1. As shown in fig. 2, the information pushing apparatus 200 includes:

the acquiring unit 201 is configured to acquire a history to be analyzed and at least two analyzed history sub-evaluation values of a target client on a target application;

the analysis unit 202 is configured to analyze the history record based on a natural language processing model to obtain at least two reference sub-evaluation values;

the updating unit 203 is configured to update the historical sub-evaluation value based on the reference sub-evaluation value to obtain a target sub-evaluation value;

the weighting unit 204 is configured to weight the target sub-evaluation value and a preset weight of a preset evaluation dimension corresponding to the target sub-evaluation value to obtain a target comprehensive evaluation value;

the selecting unit 205 is configured to select a target service from the resource pool of the target application based on the target sub-evaluation value and the target comprehensive evaluation value;

the pushing unit 206 is configured to push pushing information of the target service to the target client.

In a possible example, the analysis unit 202 is specifically configured to obtain a data type corresponding to a preset evaluation dimension based on a sprite model of the preset evaluation dimension corresponding to the historical sub-evaluation value; acquiring subdata corresponding to the data type from the historical record based on a natural language processing model; calculating the sub data based on an operational expression corresponding to the sub portrait model to obtain an evaluation value corresponding to the sub data; and weighting the evaluation value corresponding to the subdata and a preset subduer value to obtain a reference subdocument value corresponding to the preset evaluation dimension.

In a possible example, the weighting unit 204 is further configured to determine an adjustment parameter of the sub data; and adjusting the industry weight of the subdata based on the adjustment parameter to obtain the preset subdata weight.

In a possible example, the updating unit 203 is specifically configured to calculate a first ratio between a time length corresponding to the history record and a time length corresponding to the history sub-evaluation value; calculating a second ratio between the data quantity of the sub data and the data quantity corresponding to the historical sub evaluation value; a target sub-evaluation value is calculated based on the first ratio, the second ratio, the reference sub-evaluation value, and the history sub-evaluation value.

In a possible example, the selecting unit 205 is specifically configured to select, from a resource pool of the target application, a reference service corresponding to the target sub-evaluation value and the target comprehensive evaluation value; if the number of the reference services is greater than or equal to 2, determining the recommendation probability of the reference services based on the service attributes of the reference services; determining a preference value for the reference service based on the history; calculating a recommended value of the reference service based on the recommendation probability and the preference value; and selecting a target service from the reference services based on the recommended value.

In a possible example, the selecting unit 205 is specifically configured to select, from the history records, a target history record corresponding to the attribute information of the reference service; acquiring the average browsing duration and the purchasing success rate of the target customer based on the target history record; and calculating the preference value of the reference service based on the average browsing duration and the purchase success rate.

In a possible example, the pushing unit 206 is specifically configured to obtain a keyword and/or a pushed image of the target service based on the attribute information of the target service; acquiring push information of the target service based on the keyword and/or the push image; and pushing the push information to the target client.

For detailed processes executed by each unit in the information pushing apparatus 200, reference may be made to the execution steps in the foregoing method embodiments, which are not described herein again.

Referring to fig. 3, fig. 3 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure. As shown in fig. 3, the computer device 300 includes a processor 310, a memory 320, a communication interface 330, and one or more programs 340. The processor 310, the memory 320, and the communication interface 330 are interconnected via a bus 350. The related functions implemented by the obtaining unit 201, the analyzing unit 202, the updating unit 203, the weighting unit 204, the selecting unit 205 and the pushing unit 206 shown in fig. 2 can be implemented by the processor 310.

The one or more programs 340 are stored in the memory 320 and configured to be executed by the processor 310, the programs 340 including instructions for:

and pushing the pushing information of the target service to the target client.

In one possible example, in terms of analyzing the history based on the natural language processing model to obtain at least two reference sub-rating values, the program 340 is specifically configured to execute the following steps:

acquiring a data type corresponding to a preset evaluation dimension based on a subimage model of the preset evaluation dimension corresponding to the historical subinterval value;

acquiring subdata corresponding to the data type from the historical record based on a natural language processing model;

calculating the sub data based on an operational expression corresponding to the sub portrait model to obtain an evaluation value corresponding to the sub data;

and weighting the evaluation value corresponding to the subdata and a preset subduer value to obtain a reference subdocument value corresponding to the preset evaluation dimension.

In a possible example, before the weighting is performed on the evaluation value and the preset sub-weight value corresponding to the sub-data to obtain the reference sub-evaluation value corresponding to the preset evaluation dimension, the program 340 is further configured to execute the following steps:

determining an adjustment parameter of the subdata;

and adjusting the industry weight of the subdata based on the adjustment parameter to obtain the preset subdata weight.

In one possible example, in terms of updating the historical sub-evaluation value based on the reference sub-evaluation value to obtain the target sub-evaluation value, the program 340 is specifically configured to execute the following instructions:

calculating a first ratio between the time length corresponding to the historical record and the time length corresponding to the historical sub-evaluation value;

calculating a second ratio between the data quantity of the sub data and the data quantity corresponding to the historical sub evaluation value;

a target sub-evaluation value is calculated based on the first ratio, the second ratio, the reference sub-evaluation value, and the history sub-evaluation value.

In one possible example, in terms of the selecting the target service from the resource pool of the target application based on the target sub-rating value and the target comprehensive rating value, the program 340 is specifically configured to execute the following steps:

selecting reference services corresponding to the target sub-evaluation value and the target comprehensive evaluation value from a resource pool of the target application program;

if the number of the reference services is greater than or equal to 2, determining the recommendation probability of the reference services based on the service attributes of the reference services;

determining a preference value for the reference service based on the history;

calculating a recommended value of the reference service based on the recommendation probability and the preference value;

and selecting a target service from the reference services based on the recommended value.

In one possible example, in the determining the preference value for the reference service based on the history, the program 340 is specifically configured to execute the following steps:

selecting a target historical record corresponding to the attribute information of the reference service from the historical records;

acquiring the average browsing duration and the purchasing success rate of the target customer based on the target history record;

and calculating the preference value of the reference service based on the average browsing duration and the purchase success rate.

In one possible example, in terms of the pushing information of the target service to the target client, the program 340 is specifically configured to execute the following steps:

acquiring keywords and/or pushed images of the target service based on the attribute information of the target service;

acquiring push information of the target service based on the keyword and/or the push image;

and pushing the push information to the target client.

Embodiments of the present application also provide a computer storage medium, where the computer storage medium stores a computer program for causing a computer to execute to implement part or all of the steps of any one of the methods described in the method embodiments, and the computer includes an electronic device and a server.

Embodiments of the application also provide a computer program product comprising a non-transitory computer readable storage medium storing a computer program operable to cause a computer to perform to implement some or all of the steps of any of the methods recited in the method embodiments. The computer program product may be a software installation package, the computer comprising an electronic device and a server.

It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present application is not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the application. Further, those skilled in the art will also appreciate that the embodiments described in this specification are presently preferred and that no particular act or mode of operation is required in the present application.

In the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus may be implemented in other manners. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, a division of a unit is merely a logical division, and an actual implementation may have another division, for example, at least one unit or component may be combined or integrated with another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of some interfaces, devices or units, and may be an electric or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may also be distributed on at least one network unit. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a hardware mode or a software program mode.

The integrated unit, if implemented in the form of a software program module and sold or used as a stand-alone product, may be stored in a computer readable memory. With such an understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a memory and includes several instructions for causing a computer (which may be a personal computer, a server, a network device, or the like) to execute all or part of the steps of the methods according to the embodiments of the present application. And the aforementioned memory comprises: various media capable of storing program codes, such as a usb disk, a read-only memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and the like.

The foregoing detailed description of the embodiments of the present application has been presented to illustrate the principles and implementations of the present application, and the above description of the embodiments is only provided to help understand the method and the core concept of the present application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.

Claims

1. An information push method based on natural language processing is characterized by comprising the following steps:

and pushing the pushing information of the target service to the target client.

2. The method of claim 1, wherein analyzing the history based on a natural language processing model to obtain at least two reference sub-ratings comprises:

3. The method according to claim 2, wherein before the weighting the evaluation value and the preset sub-weight value corresponding to the sub-data to obtain the reference sub-evaluation value corresponding to the preset evaluation dimension, the method further comprises:

determining an adjustment parameter of the subdata;

4. The method according to claim 2, wherein the updating the historical sub-rating value based on the reference sub-rating value to obtain a target sub-rating value comprises:

5. The method according to any one of claims 1-4, wherein the selecting target traffic from the resource pool of the target application based on the target sub-rating value and the target comprehensive rating value comprises:

determining a preference value for the reference service based on the history;

6. The method of claim 5, wherein the determining the preference value for the reference service based on the history comprises:

7. The method according to any of claims 1-4, wherein said pushing information of the target service to the target client comprises:

and pushing the push information to the target client.

8. An information pushing apparatus based on natural language processing, comprising:

9. A computer device comprising a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, the programs comprising instructions for performing the steps of the method of any of claims 1-7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, the computer program causing a computer to execute to implement the method of any one of claims 1-7.