Method and device for pushing tail information

Abstract

The present invention discloses a method and device for pushing tail information, and relates to the field of computer technology. A specific implementation of the method includes: when it is necessary to push tail information to a user, a first material candidate set is determined using a pre-trained material recall model, and a second material candidate set is determined using a preset business policy; the materials in the first material candidate set and the second material candidate set are input into a pre-trained material fine-ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; the strategy weight value of each material in the second material candidate set is obtained; for each finely ranked material and / or each material in the second material candidate set, a final weight value is determined based on the model weight value and the strategy weight value of the material, and the target material to be pushed for the tail information is determined based on the final weight value. This implementation can reduce the development work of the strategy logic, and enable business personnel to quickly influence the strategy effect in real time.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for pushing tail information. Background Art

[0002] At present, as commercial banks turn more to retail and user ends, the credit card information is no longer satisfied with just prompts and notifications, and more marketing information is also placed on the end of the credit card. In order to improve the conversion rate of marketing information, it is necessary to use artificial intelligence models to sort the end of the materials. Due to the particularity of the financial industry, the end of the information needs to strictly comply with business rules, so business personnel need to use business rules to intervene in the sorting results of the model.

[0003] In actual applications, the credit card posting service contains a large number of complex business policies. For example, some preferential posting information can only be pushed to users holding a certain type of credit card, and notification and reminder posting information should be configured as the highest priority. Generally, business policies that affect the final sorting will be processed uniformly in the re-sorting stage.

[0004] Specifically, in traditional sorting services, the sorting system is divided into a recall module, a refined sorting module, and a re-sorting module. Among them, the recall module is used to select candidate sets for push from a large number of materials, which include two parts: a candidate set obtained based on business experience and a candidate set obtained through similarity calculation. After that, the material candidate set will be sent to the model refined sorting layer for sorting processing, which is completely dependent on the calculation of the model. After that, the refined sorted materials will be sent to the re-sorting module, which generates the final sorting list of materials based on the actual needs of the business and the established strategy.

[0005] The above-mentioned sorting system uses a funnel-type structure, which has the following problems: the system needs to add a lot of policy logic development work, which increases the burden on developers and makes the entire system heavy because it contains too much policy logic; the existing method makes it impossible for business personnel to change the policy through simple configuration, which reduces the speed of system response; because the time consumption of a single call is the sum of the time consumption of all modules, the time consumption requirements for all modules are relatively high, and complex business policies will greatly increase the time consumption of a single response. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides a method and device for pushing tail information, which can reduce the development work of policy logic and enable business personnel to quickly influence the policy effect in real time.

[0007] To achieve the above objective, according to one aspect of the present invention, a method for pushing tail information is provided.

[0008] The method for pushing tail information of an embodiment of the present invention includes: when it is necessary to push tail information to a user, using a pre-trained material recall model to determine a first material candidate set, and using a preset business policy to determine a second material candidate set; wherein the first material candidate set and the second material candidate set both include multiple materials; inputting the materials in the first material candidate set and the second material candidate set into a pre-trained material fine-ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; obtaining the strategy weight value of each material in the second material candidate set; for each finely ranked material and / or each material in the second material candidate set, determining a final weight value according to the model weight value and the strategy weight value of the material, and determining the target material to be pushed for the tail information based on the final weight value.

[0009] Optionally, the model weight value of each sorted material is positively correlated with the similarity between the material and the current business.

[0010] Optionally, the model weight value of each refined material is a value between zero and one; for each material that exists in the second material candidate set and does not belong to the refined material, its model weight value is zero.

[0011] Optionally, for each material in the second material candidate set: when the material needs to be placed at the top, the strategic weight value of the material is preset to a value greater than one; when the final weight value of the material needs to be increased, the strategic weight value of the material is preset to a value between zero and one.

[0012] Optionally, for the refined sorted materials that do not exist in the second material candidate set, their strategic weight values ​​are preset to zero.

[0013] Optionally, determining the final weight value according to the model weight value and the strategy weight value of the material includes: determining the sum of the model weight value and the strategy weight value of any material as the final weight value of the material.

[0014] Optionally, determining the target material for which the tail information is to be pushed according to the final weight value includes: determining the material with the largest final weight value as the target material among the refined sorted materials and / or the materials in the second material candidate set.

[0015] To achieve the above objective, according to another aspect of the present invention, a device for pushing tail information is provided.

[0016] The device for pushing tail information in an embodiment of the present invention includes: a recall unit, which is used to: when it is necessary to push tail information to a user, determine a first material candidate set by using a pre-trained material recall model, and determine a second material candidate set by using a preset business strategy; wherein the first material candidate set and the second material candidate set both include multiple materials; a fine sorting unit, which is used to input the materials in the first material candidate set and the second material candidate set into a pre-trained material fine sorting model to obtain the finely sorted materials and the model weight value of each finely sorted material; obtain the strategy weight value of each material in the second material candidate set; and a result merging unit, which is used to: for each finely sorted material and / or each material in the second material candidate set, determine a final weight value according to the model weight value and the strategy weight value of the material, and determine the target material to be pushed for tail information based on the final weight value.

[0017] Optionally, the model weight value of each refined material is a value between zero and one; for each material that exists in the second material candidate set and does not belong to the refined material, its model weight value is zero; and the refined ranking unit is further used to: for each material in the second material candidate set: when the material needs to be placed at the top, preset the strategic weight value of the material to a value greater than one; when the final weight value of the material needs to be increased, preset the strategic weight value of the material to a value between zero and one; for the refined materials that do not exist in the second material candidate set, preset their strategic weight values ​​to zero.

[0018] Optionally, the result merging unit is further used to: determine the sum of the model weight value and the strategy weight value of any material as the final weight value of the material; determine the material with the largest final weight value as the target material among the refined materials and / or materials in the second material candidate set; and the model weight value of each refined material is positively correlated with the similarity between the material and the current business.

[0019] To achieve the above objective, according to another aspect of the present invention, an electronic device is provided.

[0020] An electronic device of the present invention includes: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the tail information push method provided by the present invention.

[0021] To achieve the above objective, according to yet another aspect of the present invention, a computer-readable storage medium is provided.

[0022] A computer-readable storage medium of the present invention stores a computer program, and when the program is executed by a processor, the method for pushing tail information provided by the present invention is implemented.

[0023] According to the technical solution of the present invention, the embodiments of the above invention have the following advantages or beneficial effects: when it is necessary to push the tail information to the user, first, the first material candidate set is determined by using the pre-trained material recall model, and the second material candidate set is determined by using the preset business policy; then, the materials in the first material candidate set and the second material candidate set are input into the pre-trained material fine sorting model to obtain the finely sorted materials and the model weight value of each finely sorted material; the policy weight value of each material in the second material candidate set is obtained; thereafter, for each finely sorted material and / or each material in the second material candidate set, the final weight value is determined according to the model weight value and the policy weight value of the material, and the target material to be pushed for the tail information is determined according to the final weight value. Through the above settings, the final sorting of materials can be affected based on the policy weight value, thereby flexibly meeting the conventional policy requirements of the credit card tail scenario, greatly reducing the development work of the policy logic, and reducing the response time of the entire system.

[0024] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0026] Figure 1 Schematic diagram of the main steps of the method for pushing tail information in an embodiment of the present invention;

[0027] Figure 2 Schematic diagram of the main parts of the device for pushing tail information in an embodiment of the present invention;

[0028] Figure 3 It is a schematic diagram of components of a device for pushing tail information in an embodiment of the present invention;

[0029] Figure 4 is an exemplary system architecture diagram in which embodiments of the present invention may be applied;

[0030] Figure 5 The structure diagram of an electronic device used to implement the method for pushing tail information in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0032] It should be pointed out that the embodiments of the present invention and the technical features therein may be combined with each other without conflict.

[0033] Figure 1 1 is a schematic diagram of main steps of a method for pushing tail information according to an embodiment of the present invention.

[0034] like Figure 1 As shown, the method for pushing tail information according to the embodiment of the present invention can be specifically performed according to the following steps:

[0035] Step S101: when it is necessary to push the tail information to the user, a first material candidate set is determined by using a pre-trained material recall model, and a second material candidate set is determined by using a preset business strategy.

[0036] In an embodiment of the present invention, the tail information may be a part of the electronic information in the SMS or WeChat sent by the service provider to the user after the service is completed, and the material refers to the information object used to filter for information push, which may include notification materials, reminder materials, discount materials, etc. In this step, when it is necessary to push the tail information to the user, firstly, a large number of materials may be input into the pre-trained material recall model to determine the first material candidate set, and the preset business strategy may be used to filter the large number of materials to determine the second material candidate set. It can be understood that both the first material candidate set and the second material candidate set include multiple materials.

[0037] In actual applications, the material recall model can be a machine learning model built based on an applicable algorithm. The above business strategy can use the relevant machine learning model to screen massive materials, or it can be screened by manual judgment.

[0038] Step S102: input the materials in the first material candidate set and the second material candidate set into the pre-trained material fine ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; obtain the strategic weight value of each material in the second material candidate set.

[0039] In this step, the material ranking model can be a machine learning model built based on an applicable algorithm. After the materials in the first material candidate set and the second material candidate set are input into the material ranking model, the ranked materials and the model weight value of each ranked material can be output. Generally, the model weight value of each ranked material is positively correlated with the similarity between the material and the current business (for example, the similarity between the material and the current financial product or the similarity between the material and the current user portrait), and the model weight value of each ranked material is a value between zero and one. In practical applications, for each material that exists in the second material candidate set and does not belong to the ranked material, its model weight value can be set to zero.

[0040] In particular, the strategic weight value in this step is the weight value pre-set for each material in the second material candidate set. Specifically, for each material in the second material candidate set, when it is necessary to put the material at the top (i.e., adjust the material to the front position), the strategic weight value of the material is preset to a value greater than one; when it is necessary to increase the final weight value of the material but not put it at the top, the strategic weight value of the material is preset to a value between zero and one; when it is necessary to implement a common recall strategy (such as hot recall), the strategic weight value of the corresponding material is preset to a minimum value (for example, less than 0.01); for the refined sorted materials that do not exist in the second material candidate set, their strategic weight values ​​are preset to zero.

[0041] Step S103: for each material after refined sorting and / or each material in the second material candidate set, a final weight value is determined according to the model weight value and the strategy weight value of the material, and a target material to be pushed as the tail information is determined according to the final weight value.

[0042] In this step, the sum of the model weight value and the strategy weight value of any material can be determined as the final weight value of the material. After that, the material with the largest final weight value can be determined as the target material among the materials after fine sorting and / or the materials in the second material candidate set. In this way, business personnel can achieve flexible material sorting by configuring the strategy weight value of the material: when a material needs to be placed at the top, its strategy weight value can be set to a value greater than 1, so that its final weight value must be greater than any material with a strategy weight value of zero; when a general recall strategy needs to be implemented, the strategy weight value of each material can be set to a value below 0.01, so that the material sorting is basically determined by the model weight value; when the final weight value of a material needs to be increased but the material does not need to be placed at the top, the strategy weight value of the material can be set to a value between zero and one (for example, 0.3).

[0043] In the technical solution of the embodiment of the present invention, when it is necessary to push the tail information to the user, first, the first material candidate set is determined by using the pre-trained material recall model, and the second material candidate set is determined by using the preset business policy; then, the materials in the first material candidate set and the second material candidate set are input into the pre-trained material fine-ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; the policy weight value of each material in the second material candidate set is obtained; thereafter, for each finely ranked material and / or each material in the second material candidate set, the final weight value is determined according to the model weight value and the policy weight value of the material, and the target material to be pushed for the tail information is determined according to the final weight value. Through the above settings, the final sorting of materials can be affected based on the policy weight value, thereby flexibly meeting the conventional policy requirements of the credit card tail scenario, greatly reducing the development work of the policy logic, and reducing the response time of the entire system.

[0044] It should be noted that, for the convenience of description, the aforementioned method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the present invention is not limited by the described action sequence, and some steps can actually be performed in other sequences or simultaneously. In addition, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily necessary to implement the present invention.

[0045] In order to better implement the above-mentioned solution of the embodiment of the present invention, relevant devices for implementing the above-mentioned solution are also provided below.

[0046] See also Figure 2 As shown, the device 200 for pushing tail information provided by the embodiment of the present invention may include: a recall unit 201 , a refinement unit 202 , and a result merging unit 203 .

[0047] Among them, the recall unit 201 can be used for: when it is necessary to push the tail information to the user, using the pre-trained material recall model to determine the first material candidate set, and using the preset business strategy to determine the second material candidate set; wherein the first material candidate set and the second material candidate set both include multiple materials; the fine sorting unit 202 can be used to input the materials in the first material candidate set and the second material candidate set into the pre-trained material fine sorting model to obtain the sorted materials and the model weight value of each sorted material; obtain the strategy weight value of each material in the second material candidate set; the result merging unit 203 can be used for: for each sorted material and / or each material in the second material candidate set, determine the final weight value according to the model weight value and the strategy weight value of the material, and determine the target material as the tail information to be pushed according to the final weight value.

[0048] Preferably, in an embodiment of the present invention, the model weight value of each refined material is a value between zero and one; for each material that exists in the second material candidate set and does not belong to the refined material, its model weight value is zero; and the refined ranking unit 202 can be further used for: for each material in the second material candidate set: when it is necessary to top the material, preset the strategic weight value of the material to a value greater than one; when it is necessary to increase the final weight value of the material, preset the strategic weight value of the material to a value between zero and one; for the refined materials that do not exist in the second material candidate set, preset their strategic weight values ​​to zero.

[0049] As a preferred scheme, the result merging unit 203 can be further used to: determine the sum of the model weight value and the strategy weight value of any material as the final weight value of the material; among the refined materials and / or the materials in the second material candidate set, determine the material with the largest final weight value as the target material; and the model weight value of each refined material is positively correlated with the similarity between the material and the current business.

[0050] Figure 3 FIG. 1 is a schematic diagram of components of a device for pushing tail information in an embodiment of the present invention. Figure 3 As shown, the recall unit 201 may further include a model recall unit and a strategy recall unit, and the fine sorting unit 202 may further include a model fine sorting unit and a strategy fine sorting unit.

[0051] Among them, the model recall unit can be used to determine the first material candidate set using a pre-trained material recall model, and the strategy recall unit can be used to determine the second material candidate set using a preset business strategy. The model refinement unit is used to input the materials in the first material candidate set and the second material candidate set into the pre-trained material refinement model to obtain the refined materials and the model weight value of each refined material; the strategy refinement unit is used to obtain the strategy weight value of each material in the second material candidate set; the result merging unit is used to: for each refined material and / or each material in the second material candidate set, determine the final weight value according to the model weight value and strategy weight value of the material, and determine the target material to be pushed as the tail information based on the final weight value.

[0052] Business personnel can achieve flexible material sorting by configuring the policy weight value of the material: when a material needs to be placed at the top, its policy weight value can be set to a value greater than 1, so that its final weight value must be greater than any material with a policy weight value of zero; when a general recall strategy needs to be implemented, the policy weight value of each material can be set to a value below 0.01, so that the material sorting is basically determined by the model weight value; when the final weight value of a material needs to be increased but the material does not need to be placed at the top, the policy weight value of the material can be set to a value between zero and one (for example, 0.3). Business personnel can also achieve more complex policy requirements by modifying the corresponding rules of the policy fine sorting unit and the result merging unit. In addition, since the model recall unit, the policy recall unit, the model fine sorting unit, and the policy fine sorting unit all use multi-threading for parallel calculation, the overall time consumption of the service call is extremely limited when the number of policies increases, which can effectively reduce the response time of a single call.

[0053] In the technical solution of the embodiment of the present invention, when it is necessary to push the tail information to the user, first, the first material candidate set is determined by using the pre-trained material recall model, and the second material candidate set is determined by using the preset business policy; then, the materials in the first material candidate set and the second material candidate set are input into the pre-trained material fine-ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; the policy weight value of each material in the second material candidate set is obtained; thereafter, for each finely ranked material and / or each material in the second material candidate set, the final weight value is determined according to the model weight value and the policy weight value of the material, and the target material to be pushed for the tail information is determined according to the final weight value. Through the above settings, the final sorting of materials can be affected based on the policy weight value, thereby flexibly meeting the conventional policy requirements of the credit card tail scenario, greatly reducing the development work of the policy logic, and reducing the response time of the entire system.

[0054] Figure 4 An exemplary system architecture 400 is shown to which the method or device for pushing tail information according to the embodiment of the present invention may be applied.

[0055] like Figure 4 As shown, the system architecture 400 may include terminal devices 401, 402, 403, a network 404 and a server 405 (this architecture is only an example, and the components included in the specific architecture may be adjusted according to the specific application). The network 404 is used to provide a medium for a communication link between the terminal devices 401, 402, 403 and the server 405. The network 404 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0056] Users can use terminal devices 401, 402, 403 to interact with server 405 through network 404 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 401, 402, 403, such as information push applications (only as an example).

[0057] The terminal devices 401 , 402 , and 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0058] The server 405 may be a server that provides various services, such as a background server that provides support for information push applications operated by users using the terminal devices 401, 402, and 403 (only as an example). The background server 405 may process the received target material determination request, etc., and feed back the processing result (such as the determined target material - only as an example) to the terminal devices 401, 402, and 403.

[0059] It should be noted that the post-it information pushing method provided in the embodiment of the present invention is generally executed by the server 405 , and accordingly, the post-it information pushing device is generally arranged in the server 405 .

[0060] It should be understood that Figure 4 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0061] The present invention also provides an electronic device. The electronic device of an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method for pushing tail information provided by the present invention.

[0062] Reference below Figure 5 , which shows a schematic diagram of the structure of a computer system 500 of an electronic device suitable for implementing an embodiment of the present invention. Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0063] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the computer system 500 are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0064] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed so that a computer program read therefrom is installed into the storage section 508 as needed.

[0065] In particular, according to the embodiments disclosed in the present invention, the process described in the main step diagram above can be implemented as a computer software program. For example, the embodiments of the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the main step diagram. In the above embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the central processing unit 501, the above functions defined in the system of the present invention are executed.

[0066] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0067] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the above-mentioned module, program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0068] The units involved in the embodiments of the present invention may be implemented by software or hardware. The units described may also be set in a processor, for example, it may be described as: a processor includes a recall unit, a fine sorting unit, and a result merging unit. The names of these units do not constitute a limitation on the units themselves in some cases. For example, the recall unit may also be described as "a unit that provides a first material candidate set and a second material candidate set to a fine sorting unit."

[0069] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently and not be assembled into the device. The above computer-readable medium carries one or more programs. When the above one or more programs are executed by the device, the steps executed by the device include: when it is necessary to push tail information to the user, determine the first material candidate set using the pre-trained material recall model, and determine the second material candidate set using the preset business strategy; wherein the first material candidate set and the second material candidate set both include multiple materials; input the materials in the first material candidate set and the second material candidate set into the pre-trained material fine sorting model to obtain the finely sorted materials and the model weight value of each finely sorted material; obtain the strategy weight value of each material in the second material candidate set; for each finely sorted material and / or each material in the second material candidate set, determine the final weight value according to the model weight value and strategy weight value of the material, and determine the target material to be pushed as the tail information according to the final weight value.

[0070] In the technical solution of the embodiment of the present invention, the intervention of the tail effect can be achieved by setting the strategy weight value of the material, thereby reducing the complexity and workload of implementing complex strategies in the system, enabling business personnel to quickly influence the strategy effect in real time, while reducing the response time of a single call, leaving more execution space for the strategy.

[0071] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for pushing tail information, characterized in that: include: When it is necessary to push the tail information to the user, a first material candidate set is determined by using a pre-trained material recall model, and a second material candidate set is determined by using a preset business strategy; wherein both the first material candidate set and the second material candidate set include multiple materials; Input the materials in the first material candidate set and the second material candidate set into the pre-trained material fine ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; obtain the strategic weight value of each material in the second material candidate set; For each material after refined sorting and / or each material in the second material candidate set, a final weight value is determined according to the model weight value and the strategy weight value of the material, and a target material to be pushed tail information is determined according to the final weight value; The model weight value of each sorted material is a value between zero and one; for each material that exists in the second material candidate set and does not belong to the sorted materials, its model weight value is zero; for each material in the second material candidate set: when the material needs to be placed at the top, the strategic weight value of the material is preset to a value greater than one; when the final weight value of the material needs to be increased, the strategic weight value of the material is preset to a value between zero and one; for the sorted materials that do not exist in the second material candidate set, their strategic weight values ​​are preset to zero.

2. The method according to claim 1, characterized in that The model weight value of each refined material is positively correlated with the similarity between the material and the current business.

3. The method according to claim 1, characterized in that Determining the final weight value according to the model weight value and the strategy weight value of the material includes: The sum of the model weight value and the strategy weight value of any material is determined as the final weight value of the material.

4. The method according to claim 3, characterized in that The step of determining the target material as the target material to be pushed the tail information according to the final weight value includes: Among the refined materials and / or the materials in the second material candidate set, the material with the largest final weight value is determined as the target material.

5. A device for pushing tail information, characterized in that: include: The recall unit is used to: when it is necessary to push the tail information to the user, determine the first material candidate set by using the pre-trained material recall model, and determine the second material candidate set by using the preset business strategy; wherein both the first material candidate set and the second material candidate set include multiple materials; A fine ranking unit is used to input the materials in the first material candidate set and the second material candidate set into a pre-trained material fine ranking model to obtain the finely ranked materials and the model weight value of each finely ranked material; and obtain the strategy weight value of each material in the second material candidate set; The result merging unit is used to: determine a final weight value for each material after refined sorting and / or each material in the second material candidate set according to the model weight value and the strategy weight value of the material, and determine a target material as the target material to be pushed tail information according to the final weight value; The model weight value of each refined material is a value between zero and one; for each material that exists in the second material candidate set and does not belong to the refined material, its model weight value is zero; The fine ranking unit is further used for: for each material in the second material candidate set: when the material needs to be placed at the top, the strategic weight value of the material is preset to a value greater than one; when the final weight value of the material needs to be increased, the strategic weight value of the material is preset to a value between zero and one; for the finely ranked materials that do not exist in the second material candidate set, their strategic weight values ​​are preset to zero.

6. The device according to claim 5, characterized in that The result merging unit is further used to: The sum of the model weight value and the strategy weight value of any material is determined as the final weight value of the material; among the materials after refined sorting and / or the materials in the second material candidate set, the material with the largest final weight value is determined as the target material; as well as The model weight value of each refined material is positively correlated with the similarity between the material and the current business.

7. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 4.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images