Information recommendation system, method, and readable storage medium

Abstract

The application provides an information recommendation system, method and readable storage medium. The information recommendation system provided in the embodiments of the application is independent between modules, has a simple and clear structure, and can be used for an enterprise having a recommendation information demand to quickly build an information recommendation system of the enterprise. The information recommendation system provided in the embodiments of the application has a clear technology stack, independent processing between modules, less consumption of computing resources and storage resources, low maintenance cost, support for flexible adjustment of user recommendation data in real time, change with the change of business demand of the enterprise, and high flexibility.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of information processing, in particular to an information recommendation system, method and readable storage medium. BACKGROUND

[0002] With the development of science and technology, e-commerce and short video industry have gradually developed. In the field of e-commerce and short video, some merchants or enterprises often need to make personalized recommendations for products according to the interests of users when promoting products.

[0003] At present, in order to better promote their own products or businesses, different merchants or enterprises often deploy their own product promotion systems or business recommendation systems before the products go online. However, some enterprises have weak R&D capabilities, and the deployed recommendation system framework is messy and difficult to maintain. In the process of developing business, it may be necessary to reconstruct the framework of the existing recommendation system several times. The framework of the mature recommendation system of some enterprises is too complex, and the maintenance cost is high. The commonly used recommendation system framework has too many components and complex architecture, which not only has high maintenance difficulty, but also has poor scalability. SUMMARY

[0004] The present application aims to at least solve one of the above technical defects. Therefore, the present application provides an information recommendation system, method and readable storage medium, which can solve the technical defects of complex recommendation system architecture, difficult maintenance and poor scalability in the prior art.

[0005] An information recommendation system comprises:

[0006] a user management module, an interface module and a main logic module;

[0007] The user management module sends the business demand of the user to the interface module after receiving the business demand of the user.

[0008] The interface module receives the business demand of the user sent by the user management module, analyzes the business demand of the user, obtains the interface parameters corresponding to the business demand of the user, verifies the interface parameters corresponding to the business demand of the user, and sends the interface parameters corresponding to the business demand of the user to the main logic module if the interface parameters corresponding to the business demand of the user pass the verification.

[0009] The main logic module calls the interface corresponding to the interface parameters corresponding to the business demand of the user according to the interface parameters corresponding to the business demand of the user, obtains target data corresponding to the business demand of the user, and recommends the target data to the user management module.

[0010] Preferably, the system further comprises:

[0011] The main logic module accesses a service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user, and updates the interface parameter corresponding to the newly accessed service module to the interface module.

[0012] Preferably, the process that the main logic module accesses the service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user comprises:

[0013] The main logic module inserts a cache module according to the interface parameter corresponding to the service requirement of the user.

[0014] The main logic module calls the cache module to determine whether the service requirement of the user includes user request cache information, if it is determined that the service requirement of the user includes user request cache information, the cache module is called to cache personalized data requested by the user to be cached, if it is determined that the service requirement of the user does not include user request cache information, the cache module is called to obtain local cache data and other cache data of a task of timing update.

[0015] Preferably, the process that the main logic module accesses the service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user comprises:

[0016] The main logic module inserts a global service component according to the interface parameter corresponding to the service requirement of the user.

[0017] The global service component comprises a project universal module and a configuration module.

[0018] The main logic module calls the configuration module to configure a project file corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user, and obtains each project file corresponding to the service of the user.

[0019] The main logic module calls the project universal module to pull the project file corresponding to each service requirement of the user from each data connection pool to a project management system corresponding to the project file according to the interface parameter corresponding to the service requirement of the user, and updates the operation log of each project to a log management system corresponding to each project.

[0020] Preferably, the process that the main logic module accesses the service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user comprises:

[0021] The main logic module inserts a bottom data logic module, a cache module, an auxiliary logic module, a personalized recommendation component, a timing task management module and a file storage system according to interface parameters corresponding to service requirements of the user;

[0022] The personalized recommendation component includes a recall module, a fine arrangement module, a rearrangement module and a model file module.

[0023] The bottom data logic module sends data of an AB test switch configuration to the auxiliary logic module.

[0024] The auxiliary logic module processes the data of the AB test switch configuration to obtain a user shunt result corresponding to the data of the AB test switch configuration and sends the result to the main logic module.

[0025] The main logic module calls the cache module to determine whether there is cache data corresponding to the user shunt result according to the user shunt result.

[0026] If the cache module has cache data corresponding to the user shunt result, the first cache data corresponding to the user shunt result is sent to the main logic module.

[0027] Preferably, the system further includes:

[0028] If the cache module does not have cache data corresponding to the user shunt result, the cache data corresponding to the user shunt result is sent to the recall module.

[0029] The main logic module calls the data logic module to obtain user feature data corresponding to the user shunt result according to the user shunt result.

[0030] The main logic module calls the auxiliary logic module to send real-time inventory filtering data of the auxiliary logic module to the recall module according to the user shunt result.

[0031] The recall module processes the real-time inventory filtering data of the auxiliary logic module, the user feature data and the cache data corresponding to the user shunt result to obtain target recall data and sends the data to the main logic module.

[0032] The timing task management module regularly obtains a first target model file trained according to user behavior from the file storage system.

[0033] The main logic module receives the first target recall data and sends the data to the fine arrangement module.

[0034] The fine arrangement module receives the first target recall data, loads the first target model file corresponding to the first target recall data, and performs a prediction scoring on the first target recall data to obtain a ranking result of the first target recall data and send the ranking result to the main logic module;

[0035] The main logic module receives the ranking result of the first target recall data, analyzes the ranking result of the first target recall data to obtain a fine arrangement result of the first target recall data, and sends the fine arrangement result of the first target recall data to the rearrangement module;

[0036] The rearrangement module receives the fine arrangement result of the first target recall data, processes the fine arrangement result of the first target recall data according to a preset rearrangement strategy in cooperation with the auxiliary logic module, and finally outputs target user personalized data and sends the target user personalized data to the main logic module.

[0037] Preferably, the main logic module accesses a process of a business module corresponding to an interface parameter corresponding to the business requirement of the user according to the interface parameter corresponding to the business requirement of the user, including:

[0038] The main logic module inserts a cache module, a timing task management module, a personalized recommendation component, a file storage system, and an auxiliary logic module according to the interface parameter corresponding to the business requirement of the user;

[0039] The personalized recommendation component includes a fine arrangement module, a recall module, and a rearrangement module.

[0040] The main logic module calls the cache module to determine whether the business requirement of the user includes user request cache information, calls the cache module to cache personalized data requested by the user to be cached if it is determined that the business requirement of the user includes user request cache information, and calls the cache module to obtain local cache data of a timing update task and other cache data if it is determined that the business requirement of the user does not include user request cache information.

[0041] The recall module processes recall category data newly processed by the timing task management module every day to obtain second target recall data, and sends the second target recall data to the local cache module for storage and to the fine arrangement module;

[0042] When the business requirement of the user is a T+1 update mode, the timing task management module processes a model training task once a day according to the business requirement of the user, and stores a trained second target model file to the file storage system;

[0043] The file storage system sends the second target model file sent by the timing task management module to the fine arrangement module;

[0044] The fine arrangement module receives the second target model file, loads the second target model file corresponding to the second target recall data, and performs a prediction score on the second target recall data to obtain a ranking result of the second target recall data and send it to the main logic module;

[0045] The main logic module receives the ranking result of the second target recall data, analyzes the ranking result of the second target recall data to obtain a fine arrangement result of the second target recall data, and then transmits the fine arrangement result of the second target recall data to the rearrangement module;

[0046] The rearrangement module receives the fine arrangement result of the second target recall data, and according to a preset rearrangement strategy, cooperates with the auxiliary logic module to process the fine arrangement result of the second target recall data, and finally outputs target user personalized data and sends it to the main logic module.

[0047] Preferably, the system further comprises:

[0048] When the user's business requirement is to perform an hour-level update mode, the timing task management module processes a model training task once a day and every hour according to the user's business requirement, and stores a trained third target model file to the file storage system;

[0049] The file storage system sends the third target model file sent by the timing task management module to the fine arrangement module;

[0050] The fine arrangement module receives the third target model file, loads the third target model file corresponding to the second target recall data, and performs a prediction score on the second target recall data to obtain a ranking result of the second target recall data and send it to the main logic module;

[0051] The main logic module receives the ranking result of the second target recall data, analyzes the ranking result of the second target recall data to obtain a fine arrangement result of the second target recall data, and then transmits the fine arrangement result of the second target recall data to the rearrangement module;

[0052] The rearrangement module receives the fine arrangement result of the second target recall data, and according to a preset rearrangement strategy, cooperates with the auxiliary logic module to process the fine arrangement result of the second target recall data, and finally outputs target user personalized data and sends it to the main logic module.

[0053] Preferably, the main logic module accesses a service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user, and the process comprises:

[0054] The main logic module inserts a cache module, a timing task management module, a personalized recommendation component, a file storage system, an auxiliary logic module and a stratum data logic module according to the interface parameter corresponding to the service requirement of the user.

[0055] The personalized recommendation component comprises a precision ranking module, a recall module and a re-ranking module.

[0056] The main logic module calls the cache module to determine whether the user requests cache information is included in the service requirement of the user, and if it is determined that the user requests cache information is included in the service requirement of the user, the cache module is called to cache the personalized data requested by the user to be cached; if it is determined that the user requests cache information is not included in the service requirement of the user, the cache module is called to obtain the local cache data and other cache data of the timing update task.

[0057] The recall module processes the recall category data newly processed by the timing task management module every day to obtain third target recall data, and sends the third target recall data to the local cache module for storage and to the precision ranking module.

[0058] When the service requirement of the user is in a real-time update mode, the stratum data logic module obtains a real-time behavior sequence of the user and sends the real-time behavior sequence to the timing task management module.

[0059] The timing task management module processes a model training task in real time according to the real-time behavior sequence of the user, and stores a fourth target model file trained to the file storage system.

[0060] The file storage system sends the third target model file sent by the timing task management module to the precision ranking module.

[0061] The precision ranking module receives the fourth target model file, loads the fourth target model file corresponding to the third target recall data, and performs a prediction scoring on the third target recall data to obtain a sorting result of the third target recall data and sends the sorting result to the main logic module.

[0062] The main logic module receives the sorting result of the third target recall data, analyzes the sorting result of the third target recall data to obtain a precision ranking result of the third target recall data, and transmits the precision ranking result of the third target recall data to the re-ranking module.

[0063] The rearrangement module receives the fine arrangement result of the third target recall data, and processes the fine arrangement result of the third target recall data according to a preset rearrangement strategy in cooperation with the auxiliary logic module, and finally outputs target user personalized data and sends the target user personalized data to the main logic module.

[0064] An information recommendation method applied to the information recommendation system of any one of the preceding introduction, the method comprising:

[0065] Receiving a business requirement of a user and analyzing the business requirement of the user, determining an interface parameter corresponding to the business requirement of the user;

[0066] Verifying the interface parameter corresponding to the business requirement of the user;

[0067] If it is determined that the interface parameter corresponding to the business requirement of the user is verified, a target interface corresponding to the interface parameter corresponding to the business requirement of the user is called;

[0068] Obtaining target data corresponding to the business requirement of the user through the target interface;

[0069] Recommending the target data to the user.

[0070] Preferably, the method further comprises:

[0071] Accessing a business module corresponding to the interface parameter corresponding to the business requirement of the user according to the interface parameter corresponding to the business requirement of the user.

[0072] A readable storage medium, the readable storage medium stores computer readable instructions, the computer readable instructions are executed by one or more processors, so that one or more processors implement the steps of the information recommendation method of any one of the preceding introduction.

[0073] From the above technical solutions can be seen, when the user has an extension recommendation business or deployment of a new business recommendation module, the embodiment of the application can provide an information recommendation system, the information recommendation can include: user management module, interface module and main logic module; wherein, the user management module can receive the user's business demand after the user's business demand is sent to the interface module; the interface module can analyze the user's business demand after receiving the user's business demand sent by the user management module, get the interface parameters corresponding to the user's business demand, and verify the interface parameters corresponding to the user's business demand, if it is determined that the interface parameters corresponding to the user's business demand are verified, the interface parameters corresponding to the user's business demand are sent to the main logic module; the main logic module receives the interface parameters corresponding to the user's business demand, calls the interface corresponding to the interface parameters corresponding to the user's business demand, obtains the target data corresponding to the user's business demand and recommends the target data to the user management module.

[0074] The information recommendation system provided by the embodiment of the application is independent between each module, the structure is simple and clear, and can be used for an enterprise with recommendation information demand to quickly build its own information recommendation system. The information recommendation system technology stack provided by the embodiment of the application is clear, each module is independently processed, the consumption of computing resources and storage resources is small, the maintenance cost is low, real-time flexible adjustment of user recommendation data is supported, the system can change with the change of business demand of the enterprise, and the flexibility is high. BRIEF DESCRIPTION OF DRAWINGS

[0075] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor under the premise of the drawings.

[0076] Figure 1 The system architecture schematic diagram of an optional information recommendation system provided by the embodiment of the application;

[0077] Figure 2 The flowchart of an information recommendation method provided by the embodiment of the application;

[0078] Figure 3 The hardware structure block diagram of an information recommendation device provided by the embodiment of the application. DETAILED DESCRIPTION

[0079] With reference to the drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0080] In the existing service recommendation scheme, Flask framework is often used as a basis for research and development. Flask is a lightweight web application framework written in Python. Some recommendation schemes also use Gunicorn tools. Gunicorn is a widely used high-performance Python HTTP Server that is compatible with most web frameworks and has the characteristics of simple implementation, lightweight, high performance, etc.

[0081] Most business management systems use Docker tools. Docker is an open source application container engine that allows developers to package their applications and dependencies into a portable container that can be published to any popular Linux or Windows operating system machine. Virtualization can also be achieved. Containers are completely sandboxed and do not have any interfaces with each other.

[0082] In view of the fact that most of the current information recommendation schemes are difficult to adapt to complex and variable business needs, the present applicant has researched an information recommendation scheme. The information recommendation scheme is independent between each module, and the structure is simple and clear. The information recommendation scheme can be used by enterprises with information recommendation needs to quickly build their own information recommendation system. The technical stack of the information recommendation system provided in the embodiments of the present application is clear, each module is independently processed, the consumption of computing resources and storage resources is small, the maintenance cost is low, real-time flexible adjustment of user recommendation data is supported, and the flexibility is strong.

[0083] The method provided in the embodiments of the present application can be used in a plurality of general-purpose or special-purpose computing device environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multi-processor devices, distributed computing environments including any of the above devices or devices, and the like.

[0084] The embodiments of the present application provide an information system recommendation system and method. The information recommendation system can be applied to various information management systems, and can also be applied to various computer terminals or intelligent terminals. The execution subject can be a processor or server of a computer terminal or an intelligent terminal.

[0085] The embodiments of the present application will be described below in conjunction with Figure 1, introduce a kind of optional system architecture that can realize information recommendation given by the embodiment of the application, as shown in Figure 1 The system architecture can include: user management module, interface module and main logic module.

[0086] Among them,

[0087] The user management module can send the user's business demand to the interface module after receiving the user's business demand.

[0088] Specifically, in actual application process, the information recommendation system provided by the embodiment of the application can manage various business demands of users by using the user management module alone.

[0089] For example, the user management module can be used to manage user data, and can also be used to manage user business demand data.

[0090] The business demand can include the demand of the user requesting to recommend personalized data, and can also include the demand of the user requesting to expand a certain business module.

[0091] For example, when the user needs to deploy new business, needs to deploy new business related business module to the original system architecture, the user management module can be used to manage the business demand data required by the user.

[0092] The interface module can receive the user's business demand sent by the user management module, and after receiving the user's business demand sent by the user management module, the interface module can further analyze the user's business demand, so as to obtain the interface parameter corresponding to the user's business demand.

[0093] Specifically, the interface module can manage all interfaces of the information recommendation system.

[0094] Among them, the interface parameter corresponding to the user's business demand reflects the interface type required by the user's business and the corresponding interface ID.

[0095] Therefore, after the interface module determines the interface parameter corresponding to the user's business demand, in order to ensure data security and system stability, it is necessary to further verify the interface parameter corresponding to the user's business demand.

[0096] Therefore, in addition to determining the interface parameter corresponding to the user's business demand, the interface module in the information recommendation system provided by the embodiment of the application can further verify the interface parameter corresponding to the user's business demand, to verify whether the interface parameter corresponding to the user's business demand is the interface parameter truly matched with the user's business demand.

[0097] In actual application process, since the parameter check is the parameter rule agreed by the front and back end when designing the interface, the security can be considered, and the user id and other key parameters can be encoded to generate the sign signature according to the agreed rule, and the check process can be designed to add a decorator method to the interface method to check each parameter according to the rule.

[0098] If the interface module determines that the interface parameter corresponding to the business demand of the user passes the check, the interface module can further send the interface parameter corresponding to the business demand of the user to the main logic module, so that the main logic module further processes the business demand of the user.

[0099] The main logic module can receive the interface parameter corresponding to the business demand of the user generated by the interface module.

[0100] Specifically, after receiving the interface parameter corresponding to the business demand of the user generated by the interface module, the main logic module can call the interface corresponding to the interface parameter corresponding to the business demand of the user according to the interface parameter corresponding to the business demand of the user, and can obtain the target data corresponding to the business demand of the user and recommend the target data to the user management module.

[0101] As can be seen from the above technical solutions, the information recommendation system provided by the embodiments of the present application is independent between each module, the structure is simple and clear, and can be used to quickly build the information recommendation system of the enterprise with recommendation information demand. The technical stack of the information recommendation system provided by the embodiments of the present application is clear, each module is independently processed, the consumption of computing resources and storage resources is small, the maintenance cost is low, real-time flexible adjustment of user recommendation data is supported, the system can change with the change of business demand of the enterprise, and the flexibility is strong.

[0102] Further optionally, as can be seen from the above introduction, in the information management system provided by the embodiments of the present application, the user management module can manage the data related to the demand of the user.

[0103] In actual application process, when the interface module determines the interface parameter corresponding to the business demand of the user, if it is found that the interface data corresponding to the interface parameter corresponding to the business demand of the user does not exist in the interface module of the information recommendation system provided by the embodiments of the present application, the main logic module of the information recommendation system provided by the embodiments of the present application can further access the business module corresponding to the interface parameter corresponding to the business demand of the user according to the interface parameter corresponding to the business demand of the user.

[0104] For example, when a user needs to add a service module of a new service, and the interface module does not have an interface of a service module corresponding to the new service of the user, the main logic module of the information recommendation system provided in the embodiment of the present application can access a service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user, and update the interface parameter corresponding to the newly accessed service module to the interface module.

[0105] As can be seen from the technical solutions described above, the information recommendation system provided in the embodiment of the present application has independent modules, simple and clear structure, and can be used by enterprises having recommendation information requirements to quickly build their own information recommendation systems. In addition, the information recommendation system can access service modules related to the service requirements of users according to different service requirements of users, which helps different users to expand service modules according to their different service requirements, perfect and deploy their own information recommendation system framework. The information recommendation system provided in the embodiment of the present application has clear technical stack, and the modules are processed independently, which can reduce the consumption of computing resources and storage resources, lower the maintenance cost, support real-time flexible adjustment of recommendation data for users, and change with the changes of service requirements of enterprises, and has high flexibility.

[0106] In actual application, the process of accessing a service module corresponding to the interface parameter corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user by the main logic module will also be different according to different service requirements of users. Next, several optional implementation manners are introduced, which can include the following.

[0107] The first,

[0108] After determining the interface parameter corresponding to the service requirement of the user, if it is determined that the interface parameter corresponding to the service requirement of the user is the interface parameter corresponding to the cache module.

[0109] The main logic module can insert the cache module according to the interface parameter corresponding to the service requirement of the user.

[0110] As described above, the user management module can send the service requirement of the user to the interface module.

[0111] After the interface module analyzes the service requirement of the user and obtains the interface parameter corresponding to the service requirement of the user, the interface module sends the interface parameter corresponding to the service requirement of the user to the main logic module after verifying the interface parameter.

[0112] Therefore, the interface parameter corresponding to the service requirement of the user can feed back the related information of the service requirement of the user.

[0113] In practical application, the cache module can cache personalized data corresponding to the service requirement of the user, and can also obtain locally cached data and other cached data of a task updated at a fixed time.

[0114] Therefore, after the cache module is inserted, the main logic module can determine whether the service requirement of the user includes a request for cached information, and if it is determined that the service requirement of the user includes a request for cached information, the cache module can be called to cache personalized data requested by the user.

[0115] If it is determined that the service requirement of the user does not include a request for cached information, the cache module can be called to obtain locally cached data and other cached data of a task updated at a fixed time.

[0116] Secondly,

[0117] After determining the interface parameter corresponding to the service requirement of the user, if it is determined that the interface parameter corresponding to the service requirement of the user is an interface parameter corresponding to a global service component.

[0118] The main logic module can insert the global service component according to the interface parameter corresponding to the service requirement of the user.

[0119] The global service component can include a project universal module and a configuration module.

[0120] As can be seen from the above description, the user management module can send the service requirement of the user to the interface module.

[0121] After the interface module analyzes the service requirement of the user and obtains the interface parameter corresponding to the service requirement of the user, the interface parameter corresponding to the service requirement of the user is sent to the main logic module after being verified.

[0122] Therefore, the interface parameter corresponding to the service requirement of the user can feed back the relevant information of the service requirement of the user.

[0123] In practical application, the service requirement of the user can include configuring a related file of a certain project or deploying a new project.

[0124] Therefore, when it is necessary to deploy a project related file corresponding to the requirement of the user, the main logic module can configure the project file corresponding to the service requirement of the user according to the interface parameter corresponding to the service requirement of the user, so as to obtain each project file corresponding to the service requirement of the user.

[0125] The main logic module can also call the project general module to pull the project file corresponding to the business requirement of each user from each data connection pool to the project management system corresponding to each project file according to the interface parameter corresponding to the business requirement of the user, and can also update the operation log of each project to the log management system corresponding to each project.

[0126] The third,

[0127] As can be seen from the above introduction, the interface parameter corresponding to the business requirement of the user can feed back the relevant information of the business requirement of the user.

[0128] Therefore, after determining the interface parameter corresponding to the business requirement of the user, if it is determined that the interface parameter corresponding to the business requirement of the user is the interface parameter corresponding to the underlying data logic module, the cache module, the auxiliary logic module, the personalized recommendation component, the timing task management module and the file storage system.

[0129] The main logic module inserts the underlying data logic module, the cache module, the auxiliary logic module, the personalized recommendation component, the timing task management module and the file storage system according to the interface parameter corresponding to the business requirement of the user.

[0130] The personalized recommendation component can include the recall module, the precision sorting module, the rearrangement module and the first target model file module.

[0131] Among them,

[0132] In actual application process, when the user's demand is to carry on AB precision sorting model test, the underlying data logic module can obtain the data of AB test switch configuration and send to the auxiliary logic module.

[0133] The auxiliary logic module can process the data of the AB test switch configuration after receiving the data of the AB test switch configuration, obtain the user's shunt result corresponding to the data of the AB test switch configuration and send to the main logic module.

[0134] As can be seen from the above introduction, the cache module can determine whether there is personalized information corresponding to the business requirement of the user.

[0135] Therefore, after receiving the user's shunt result corresponding to the data of the AB test switch configuration, the main logic module can further call the cache module to determine whether there is cache data corresponding to the user's shunt result according to the user's shunt result.

[0136] If the cache module determines that there is cached data corresponding to the user's diversion result, the cache module can send the first cached data corresponding to the user's diversion result to the main logic module, so that the main logic module can process the first cached data corresponding to the user's diversion result.

[0137] If the cache module determines that there is no cached data corresponding to the user's diversion result, the cached data corresponding to the user's diversion result is sent to the recall module, so that the recall module processes the cached data corresponding to the user's diversion result.

[0138] The main logic module can also call the data logic module to obtain user feature data corresponding to the user's diversion result according to the user's diversion result.

[0139] The main logic module can also call the auxiliary logic module to send real-time inventory filtering data of the auxiliary logic module to the recall module according to the user's diversion result, so that the recall module processes the real-time inventory filtering data of the auxiliary logic module.

[0140] After receiving the real-time inventory filtering data of the auxiliary logic module, the user feature data, and the cached data corresponding to the user's diversion result, the recall module can process the real-time inventory filtering data of the auxiliary logic module, the user feature data, and the cached data corresponding to the user's diversion result, obtain target recall data, and send it to the main logic module, so that the main logic module processes the first target recall data.

[0141] The timing task management module can periodically obtain model files trained according to user behavior from the file storage system.

[0142] The file storage system stores at least one model file corresponding to user behavior.

[0143] The first target model file can be a model file obtained by the file storage system training according to the behavior characteristics of each user as training data.

[0144] The file storage system stores at least one model file corresponding to user behavior.

[0145] The model file is a certain type of prediction algorithm model selected by an algorithm engineer according to a scene, and the algorithm model file is pre-trained offline with user portraits, user historical behavior, etc. as training samples.

[0146] The model file mainly contains some index parameters trained by the algorithm. Online, the corresponding algorithm package is used to load the model file to predict the recall data of the user, and the prediction result is sorted.

[0147] Each of the first target model files has a corresponding relationship with the first target recall data.

[0148] The main logic can receive the first target recall data and send it to the fine ranking module for processing.

[0149] The fine ranking module can receive the first target recall data and load the first target model file corresponding to the first target recall data after receiving the first target recall data.

[0150] After loading the first target model file corresponding to the first target recall data, the first target recall data can be predicted and scored to obtain the sorting result of the first target recall data and send it to the main logic module for processing.

[0151] The main logic module can receive the sorting result of the first target recall data and analyze the sorting result of the first target recall data after receiving the sorting result corresponding to the first target recall data. The fine ranking result of the first target recall data can be obtained.

[0152] The fine ranking result of the first target recall data is obtained according to the scoring prediction result of the first target recall data, so the fine ranking result of the first target recall data can match the business needs of the user. Therefore, after obtaining the fine ranking result of the first target recall data, the fine ranking result of the first target recall data can be transmitted to the rearrangement module for rearrangement processing.

[0153] The rearrangement module can receive the fine ranking result of the first target recall data and process the fine ranking result of the first target recall data according to the preset rearrangement strategy and the auxiliary logic module after receiving the fine ranking result of the first target recall data. Finally, the target user personalized data is output and sent to the main logic module, so that the main logic module can process the target user personalized data.

[0154] The fourth,

[0155] After determining the interface parameter corresponding to the service requirement of the user, if it is determined that the interface parameter corresponding to the service requirement of the user is the interface parameter corresponding to the cache module, the timing task management module, the personalized recommendation component, the file storage system and the auxiliary logic module.

[0156] Then the main logic module can insert the cache module, the timing task management module, the personalized recommendation component, the file storage system and the auxiliary logic module according to the interface parameter corresponding to the service requirement of the user.

[0157] The personalized recommendation component can include a precision sorting module, a recall module and a reordering module.

[0158] As can be seen from the above introduction, the user management module can send the service requirement of the user to the interface module.

[0159] After the interface module analyzes the service requirement of the user and obtains the interface parameter corresponding to the service requirement of the user, the interface module sends the interface parameter corresponding to the service requirement of the user to the main logic module after verification.

[0160] Therefore, the interface parameter corresponding to the service requirement of the user can feed back the related information of the service requirement of the user.

[0161] In actual application process, when the user needs to build the framework of the timing update service;

[0162] The main logic module can call the cache module to determine whether the service requirement of the user includes user request cache information, if it is determined that the service requirement of the user includes user request cache information, the cache module can be called to cache personalized data requested by the user to be cached; if it is determined that the service requirement of the user does not include user request cache information, the cache module can be called to obtain local cache data and other cache data of the timing update task.

[0163] The recall module can process the recall category data processed by the timing task management module every day to obtain second target recall data, and send the second target recall data to the local cache module for storage and to the precision sorting module;

[0164] When the service requirement of the user is the T+1 update mode, the timing task management module can process the model training task once a day according to the service requirement of the user, and store the trained second target model file to the file storage system;

[0165] The file storage system can send the first target model file sent by the timing task management module to the precision sorting module.

[0166] The fine arrangement module can receive the second target model file, load the second target model file corresponding to the second target recall data, and perform prediction scoring on the second target recall data to obtain a ranking result of the second target recall data and send the ranking result to the main logic module;

[0167] The main logic module can receive the ranking result of the second target recall data, analyze the ranking result of the second target recall data to obtain a fine arrangement result of the second target recall data, and send the fine arrangement result of the second target recall data to the rearrangement module;

[0168] The rearrangement module can receive the fine arrangement result of the second target recall data, and perform processing on the fine arrangement result of the second target recall data according to a preset rearrangement strategy in cooperation with the auxiliary logic module, and finally output target user personalized data and send the target user personalized data to the main logic module.

[0169] When the business requirement of the user is the hour-level update mode, the timing task management module can process a model training task once a day and once an hour according to the business requirement of the user, and store a trained third target model file to the file storage system;

[0170] The file storage system can send the second target model file sent by the timing task management module to the fine arrangement module;

[0171] The fine arrangement module can receive the third target model file, load the third target model file corresponding to the second target recall data, and perform prediction scoring on the second target recall data to obtain a ranking result of the second target recall data and send the ranking result to the main logic module;

[0172] The main logic module can receive the ranking result of the second target recall data, analyze the ranking result of the second target recall data to obtain a fine arrangement result of the second target recall data, and send the fine arrangement result of the second target recall data to the rearrangement module;

[0173] The rearrangement module can receive the fine arrangement result of the second target recall data, and perform processing on the fine arrangement result of the second target recall data according to a preset rearrangement strategy in cooperation with the auxiliary logic module, and finally output target user personalized data and send the target user personalized data to the main logic module.

[0174] Fifthly,

[0175] After determining the interface parameter corresponding to the service requirement of the user, if it is determined that the interface parameter corresponding to the service requirement of the user is the interface parameter corresponding to the cache module, the timing task management module, the personalized recommendation component, the file storage system, the auxiliary logic module and the stratum data logic module.

[0176] Then the main logic module can insert the cache module, the timing task management module, the personalized recommendation component, the file storage system, the auxiliary logic module and the stratum data logic module according to the interface parameter corresponding to the service requirement of the user.

[0177] The personalized recommendation component comprises a precision ranking module, a recall module and a re-ranking module.

[0178] As can be seen from the above introduction, the user management module can send the service requirement of the user to the interface module.

[0179] After the interface module analyzes the service requirement of the user and obtains the interface parameter corresponding to the service requirement of the user, the interface module sends the interface parameter corresponding to the service requirement of the user to the main logic module after verification.

[0180] Therefore, the interface parameter corresponding to the service requirement of the user can feed back the related information of the service requirement of the user.

[0181] In actual application process, when the service requirement of the user is to build an hourly level update framework.

[0182] The main logic module can insert the cache module, the timing task management module, the personalized recommendation component, the file storage system, the auxiliary logic module and the stratum data logic module according to the interface parameter corresponding to the service requirement of the user.

[0183] The main logic module can call the cache module to determine whether the service requirement of the user includes user request cache information, if it is determined that the service requirement of the user includes user request cache information, the cache module can be called to cache personalized data requested by the user to be cached; if it is determined that the service requirement of the user does not include user request cache information, the cache module can be called to obtain local cache data and other cache data of the timing update task.

[0184] The recall module can process the recall category data newly processed by the timing task management module every day to obtain third target recall data, and send the third target recall data to the local cache module for storage and to the precision ranking module.

[0185] When the service requirement of the user is to perform real-time updating mode, the stratum data logic module can acquire the real-time behavior sequence of the user and send it to the timing task management module;

[0186] The timing task management module can process the model training task in real time according to the real-time behavior sequence of the user, and can store the trained fourth target model file to the file storage system;

[0187] The file storage system can send the fourth target model file sent by the timing task management module to the fine arrangement module;

[0188] The fine arrangement module can receive the fourth target model file, load the fourth target model file corresponding to the third target recall data, perform prediction scoring on the third target recall data, obtain the sorting result of the third target recall data, and send it to the main logic module;

[0189] The main logic module can receive the sorting result of the third target recall data, analyze the sorting result of the third target recall data to obtain the fine arrangement result of the third target recall data, and then send the fine arrangement result of the third target recall data to the rearrangement module;

[0190] The rearrangement module can receive the fine arrangement result of the third target recall data, process the fine arrangement result of the third target recall data according to a preset rearrangement strategy, cooperate with the auxiliary logic module, and finally output the target user personalized data and send it to the main logic module.

[0191] As can be seen from the above technical solutions, the information recommendation system provided by the embodiments of the present application is independent between each module, the structure is simple and clear, and can be used to quickly build the information recommendation system of the enterprise with recommendation information demand. In addition, the business module related to the user's demand can be accessed according to the different business requirements of the user, which helps different users to expand the business module according to their different business, perfect and deploy their own information recommendation system framework. The information recommendation system technology stack provided by the embodiments of the present application is clear, and each module is independently processed, which can reduce the consumption of computing resources and storage resources, reduce the maintenance cost, support real-time flexible adjustment of user recommendation data, and has strong flexibility.

[0192] The information recommendation method provided by the embodiments of the present application will be described below. The information recommendation method described below can be correspondingly referred to the information recommendation system described above.

[0193] The information recommendation method provided by the embodiments of the present application will be described below. The information recommendation method described below can be correspondingly referred to the information recommendation system described above. Figure 2 , the flow of the information recommendation method given by the embodiments of the present application is introduced, as followsFigure 2 As shown in the figure, the flow can include the following steps:

[0194] Step S101, receiving a service requirement of a user and analyzing the service requirement of the user to determine an interface parameter corresponding to the service requirement of the user.

[0195] Specifically, as introduced above, the information recommendation system provided by the embodiments of the present application includes the user management module, the interface module and the main logic module. Among them, each module is independent of each other and can cooperate with each other to complete the information recommendation process corresponding to the service requirement of the user.

[0196] Therefore, the method provided by the embodiments of the present application can receive a service requirement of a user and analyze the service requirement of the user to determine an interface parameter corresponding to the service requirement of the user.

[0197] Among them, the service requirement of the user can include a request for recommended information and a corresponding interface parameter.

[0198] For example, the service requirement of the user can include that the user's business may need to access which type of interface and the corresponding interface ID.

[0199] Therefore, by analyzing the service requirement of the user, the interface parameter corresponding to the service requirement of the user can be determined.

[0200] Step S102, verifying the interface parameter corresponding to the service requirement of the user.

[0201] Specifically, as introduced above, the method provided by the embodiments of the present application can determine an interface parameter corresponding to the service requirement of the user.

[0202] In order to ensure the security of the information recommendation system and the accuracy of the data, after receiving the interface parameter corresponding to the service requirement of the user, the interface parameter corresponding to the service requirement of the user can be verified.

[0203] If it is determined that the interface parameter corresponding to the service requirement of the user is verified, it means that this time the user's request is a qualified request, and calling the target interface corresponding to the interface parameter corresponding to the service requirement of the user will not endanger the security of the information system, then step S103 can be executed.

[0204] Step S103, if it is determined that the interface parameter corresponding to the service requirement of the user is verified, the target interface corresponding to the interface parameter corresponding to the service requirement of the user is called.

[0205] Specifically, as introduced above, the method provided in the embodiments of the present application can verify the interface parameters corresponding to the service requirement of the user. It can be determined whether the verification of the interface parameters corresponding to the service requirement of the user is passed. If it is determined that the verification of the interface parameters corresponding to the service requirement of the user is passed, it means that the request of the user is a qualified request, and calling the target interface corresponding to the interface parameters corresponding to the service requirement of the user will not endanger the security of the information system. Therefore, the target interface corresponding to the interface parameters corresponding to the service requirement of the user can be called, so that the target data corresponding to the service requirement of the user can be obtained by accessing the target interface corresponding to the interface parameters corresponding to the service requirement of the user.

[0206] In step S104, the target data corresponding to the service requirement of the user is obtained through the target interface.

[0207] Specifically, as introduced above, the method provided in the embodiments of the present application can determine whether the verification of the interface parameters corresponding to the service requirement of the user is passed, and can call the target interface corresponding to the interface parameters corresponding to the service requirement of the user. The target interface and the service requirement of the user correspond to each other, which means that the relevant data corresponding to the service requirement of the user can be determined by accessing the target interface. Therefore, after the target interface is called, the target data corresponding to the service requirement of the user can be obtained by accessing the target interface.

[0208] In step S105, the target data is recommended to the user.

[0209] Specifically, as introduced above, the method provided in the embodiments of the present application can determine the target data, and the target data and the service requirement of the user correspond to each other. Therefore, the target data is the target data required by the user to carry out the business. After the target data is determined, the target data can be further recommended to the user.

[0210] As can be seen from the technical solutions introduced above, the information recommendation method provided in the embodiments of the present application can recommend the target data corresponding to the service requirement of the user to the user according to different service requirements of the user. The information recommendation method provided in the embodiments of the present application has clear logic, and can be used by enterprises with the demand of recommending information to quickly build their own information recommendation systems. It is helpful for different users to improve and deploy their own information recommendation system framework according to their own different businesses. In addition, the method provided in the embodiments of the present application has simple logic architecture, can consume less computing resources and storage resources, has low maintenance cost, supports flexible adjustment of recommended data of the user in real time, and can change with the change of the business requirement of the enterprise, and has strong flexibility.

[0211] ​​​​​​In actual application process, the information recommendation method provided by the embodiment of the application can also expand a new service recommendation module according to the demand of the user, and be used for recommending a new service corresponding to the demand of the user. Next, the process is introduced, which can include the following:

[0212] As can be known from the above introduction, the method provided by the embodiment of the application can determine the interface parameter corresponding to the service demand of the user. If it is found that the interface corresponding to the interface parameter corresponding to the service demand of the user is not included in the existing module, the method provided by the embodiment of the application can also access the service module corresponding to the interface parameter corresponding to the service demand of the user according to the interface parameter corresponding to the service demand of the user.

[0213] As can be known from the above introduction, the method provided by the embodiment of the application can determine the interface parameter corresponding to the service demand of the user. If it is found that the interface corresponding to the interface parameter corresponding to the service demand of the user is not included in the existing module, the method provided by the embodiment of the application can also access the service module corresponding to the interface parameter corresponding to the service demand of the user according to the interface parameter corresponding to the service demand of the user.

[0214] The information recommendation method provided by the embodiment of the application can be applied to an information recommendation device, such as a terminal: a mobile phone, a computer, etc. Optionally, Figure 3 The hardware structure block diagram of the information recommendation device is shown, and the hardware structure of the information recommendation device can include at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4. Figure 3

[0215] In the embodiment of the application, the number of the processor 1, the communication interface 2, the memory 3 and the communication bus 4 is at least one, and the processor 1, the communication interface 2 and the memory 3 complete the communication among each other through the communication bus 4.

[0216] The processor 1 can be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiment of the application, etc.

[0217] The memory 3 can include a high-speed RAM memory, and can also include a non-volatile memory, etc., for example, at least one disk memory.

[0218] ​The memory stores a program, and the processor can invoke the program stored in the memory, and the program is used for implementing each processing flow in the terminal information recommendation scheme.

[0219] The application further provides a readable storage medium, which can store a program suitable for processor execution, and the program is used for implementing each processing flow in the terminal information recommendation scheme.

[0220] Finally, it should be noted that, in this document, the relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0221] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between various embodiments can be referred to each other.

[0222] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the application. The various embodiments can be combined with each other. Therefore, the application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An information recommendation system, characterized in that, include: User management module, interface module, and main logic module; After receiving the user's business requirements, the user management module sends the user's business requirements to the interface module. The interface module receives the user's business requirements sent by the user management module, analyzes the user's business requirements, obtains the interface parameters corresponding to the user's business requirements, verifies the interface parameters corresponding to the user's business requirements, and if the interface parameters corresponding to the user's business requirements pass the verification, it sends the interface parameters corresponding to the user's business requirements to the main logic module. The main logic module calls the interface corresponding to the interface parameters corresponding to the user's business needs based on the interface parameters corresponding to the user's business needs, obtains the target data corresponding to the user's business needs, and recommends the target data to the user management module. The main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business needs based on the interface parameters corresponding to the user's business needs, and updates the interface parameters corresponding to the newly accessed business module to the interface module. The process by which the main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business needs, based on the interface parameters corresponding to the user's business needs, includes: the main logic module inserting a cache module, a scheduled task management module, a personalized recommendation component, a file storage system, and an auxiliary logic module based on the interface parameters corresponding to the user's business needs; wherein, the personalized recommendation component includes a fine-ranking module, a recall module, and a re-ranking module; the main logic module calls the cache module to determine whether the user's business needs include user request cache information; if it is determined that the user's business needs include user request cache information, then the cache module is called to cache the personalized data of the user request cache; if it is determined that the user's business needs do not include user request cache information, then the cache module is called to obtain the local cache data of the scheduled updated task and other cache data; the recall module processes the recall category data that is newly processed by the scheduled task management module every day to obtain the second target recall data, and sends it to the local cache module for storage and to the fine-ranking module; when the user's business needs are to perform... In T+1 update mode, the scheduled task management module processes the model training task once a day according to the user's business needs and stores the trained second target model file in the file storage system. The file storage system sends the second target model file sent by the scheduled task management module to the fine ranking module. The fine ranking module receives the second target model file, loads the second target model file corresponding to the second target recall data, performs prediction scoring on the second target recall data, obtains the ranking result of the second target recall data, and sends it to the main logic module. The main logic module receives the ranking result of the second target recall data, analyzes the ranking result of the second target recall data to obtain the fine ranking result of the second target recall data, and then passes the fine ranking result of the second target recall data to the re-ranking module. The re-ranking module receives the fine ranking result of the second target recall data, and processes the fine ranking result of the second target recall data in conjunction with the auxiliary logic module according to the preset re-ranking strategy, and finally outputs the target user personalized data and sends it to the main logic module.

2. The system according to claim 1, characterized in that, The process by which the main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business requirements, based on the interface parameters corresponding to the user's business requirements, includes: The main logic module inserts a cache module based on the interface parameters corresponding to the user's business requirements; The main logic module calls the cache module to determine whether the user's business requirements include user request cache information. If it is determined that the user's business requirements include user request cache information, the cache module is called to cache the personalized data of the user request cache. If it is determined that the user's business requirements do not include user request cache information, the cache module is called to obtain the local cache data of the periodically updated task and other cache data.

3. The system according to claim 1, characterized in that, The process by which the main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business requirements, based on the interface parameters corresponding to the user's business requirements, includes: The main logic module inserts global service components based on the interface parameters corresponding to the user's business requirements; The global service component includes a general project module and a configuration module; The main logic module calls the configuration module to configure the project files corresponding to the user's business requirements based on the interface parameters corresponding to the user's business requirements, thereby obtaining various project files corresponding to the user's business. The main logic module, based on the interface parameters corresponding to the user's business requirements, calls the project general module to pull the project files corresponding to each user's business requirements from each data connection pool to the project management system corresponding to each project file, and updates the operation logs of each project to the log management system corresponding to each project.

4. The system according to claim 1, characterized in that, The process by which the main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business requirements, based on the interface parameters corresponding to the user's business requirements, includes: The main logic module inserts an underlying data logic module, a cache module, an auxiliary logic module, a personalized recommendation component, a scheduled task management module, and a file storage system based on the interface parameters corresponding to the user's business needs. The personalized recommendation component includes a recall module, a fine ranking module, a re-ranking module, and a model file module. The underlying data logic module obtains the AB test switch configuration data and sends it to the auxiliary logic module; The auxiliary logic module processes the data configured by the AB test switch to obtain the user's traffic distribution result corresponding to the data configured by the AB test switch and sends it to the main logic module. The main logic module calls the cache module to determine whether there is cached data corresponding to the user's traffic distribution result based on the user's traffic distribution result; If the caching module has cached data corresponding to the user's traffic splitting result, then the first cached data corresponding to the user's traffic splitting result is sent to the main logic module.

5. The system according to claim 4, characterized in that, The system also includes: If the caching module does not have cached data corresponding to the user's traffic splitting result, then the cached data corresponding to the user's traffic splitting result will be sent to the recall module; The main logic module calls the data logic module to obtain user feature data corresponding to the user's traffic distribution result based on the user's traffic distribution result; Based on the user's traffic distribution result, the main logic module calls the auxiliary logic module to send the real-time inventory filtering data of the auxiliary logic module to the recall module; The recall module processes the real-time inventory filtering data of the auxiliary logic module, the user feature data, and the cached data corresponding to the user's traffic distribution result to obtain target recall data and send it to the main logic module. The scheduled task management module periodically retrieves the first target model file trained based on user behavior from the file storage system; The main logic module receives the first target recall data and sends the first target recall data to the fine ranking module; The fine ranking module receives the first target recall data, loads the first target model file corresponding to the first target recall data, performs prediction scoring on the first target recall data, obtains the ranking result of the first target recall data, and sends it to the main logic module. The main logic module receives the sorting result of the first target recall data, analyzes the sorting result of the first target recall data to obtain the fine ranking result of the first target recall data, and then transmits the fine ranking result of the first target recall data to the rearrangement module. The reordering module receives the fine-ranking result of the first target recall data, and processes the fine-ranking result of the first target recall data in conjunction with the auxiliary logic module according to the preset reordering strategy, and finally outputs the target user personalized data and sends it to the main logic module.

6. The system according to claim 1, characterized in that, The system also includes: When the user's business requirement is to perform hourly updates, the scheduled task management module processes the model training task once a day and once an hour according to the user's business requirement, and stores the trained third target model file in the file storage system. The file storage system sends the third target model file sent by the scheduled task management module to the fine sorting module; The fine ranking module receives the third target model file, loads the third target model file corresponding to the second target recall data, performs prediction scoring on the second target recall data, obtains the ranking result of the second target recall data, and sends it to the main logic module. The main logic module receives the sorting result of the second target recall data, analyzes the sorting result of the second target recall data to obtain the fine ranking result of the second target recall data, and then transmits the fine ranking result of the second target recall data to the rearrangement module. The reordering module receives the fine-ranking result of the second target recall data, and processes the fine-ranking result of the second target recall data in conjunction with the auxiliary logic module according to the preset reordering strategy, and finally outputs the target user personalized data and sends it to the main logic module.

7. The system according to claim 1, characterized in that, The process by which the main logic module accesses the business module corresponding to the interface parameters corresponding to the user's business requirements, based on the interface parameters corresponding to the user's business requirements, includes: The main logic module inserts a cache module, a scheduled task management module, a personalized recommendation component, a file storage system, an auxiliary logic module, and a stratum data logic module based on the interface parameters corresponding to the user's business needs. The personalized recommendation component includes a fine-ranking module, a recall module, and a rearrangement module. The main logic module calls the cache module to determine whether the user's business requirements include user request cache information. If it is determined that the user's business requirements include user request cache information, the cache module is called to cache the personalized data of the user request cache. If it is determined that the user's business requirements do not include user request cache information, the cache module is called to obtain the local cache data of the periodically updated task and other cache data. The recall module processes the recall category data that is newly processed by the scheduled task management module every day to obtain the third target recall data, and sends it to the local cache module for storage and to the fine ranking module. When the user's business requirement is to perform a real-time update, the stratum data logic module obtains the user's real-time behavior sequence and sends it to the scheduled task management module; The scheduled task management module processes the model training task in real time based on the user's real-time behavior sequence and stores the trained fourth target model file in the file storage system. The file storage system sends the third target model file sent by the scheduled task management module to the fine sorting module; The fine ranking module receives the fourth target model file, loads the fourth target model file corresponding to the third target recall data, performs prediction scoring on the third target recall data, obtains the ranking result of the third target recall data, and sends it to the main logic module. The main logic module receives the sorting result of the third target recall data, analyzes the sorting result of the third target recall data to obtain the fine ranking result of the third target recall data, and then transmits the fine ranking result of the third target recall data to the rearrangement module. The reordering module receives the fine-ranking result of the third target recall data, and processes the fine-ranking result of the third target recall data in conjunction with the auxiliary logic module according to the preset reordering strategy, and finally outputs the target user personalized data and sends it to the main logic module.

8. An information recommendation method, characterized in that, Applied to the information recommendation system according to any one of claims 1-7, the method comprises: Receive and analyze user business requirements, and determine the interface parameters corresponding to the user's business requirements; Validate the interface parameters corresponding to the user's business requirements; If the interface parameters corresponding to the user's business requirements are verified as valid, then the target interface corresponding to the interface parameters corresponding to the user's business requirements is invoked. The target data corresponding to the user's business needs is obtained through the target interface; The target data is recommended to the user.

9. The method according to claim 8, characterized in that, The method also includes: Based on the interface parameters corresponding to the user's business requirements, access the business module corresponding to the interface parameters corresponding to the user's business requirements.

10. A readable storage medium, characterized in that: The readable storage medium stores computer-readable instructions that, when executed by one or more processors, cause the one or more processors to implement the steps of the information recommendation method as described in any one of claims 8 to 9.

Images