CN116523587A

CN116523587A - Data processing method and related device

Info

Publication number: CN116523587A
Application number: CN202310359713.6A
Authority: CN
Inventors: 郭威; 孟昶; 张恒煜; 郭慧丰; 唐睿明
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2023-03-31
Filing date: 2023-03-31
Publication date: 2023-08-01

Abstract

The information data processing method can be applied to the field of information recommendation of artificial intelligence, and comprises the following steps: acquiring a first feature representation and a second feature representation; the first characteristic is expressed as being obtained based on first operation information of the user on the article, the first operation information indicates whether the first operation behavior of the user on the article exists or not, and the second characteristic is expressed as being obtained based on second operation information of the user on the article; the second operation information indicates whether the user has a second operation action on the article or not, and the first operation action is the association action on the article before the user performs the second operation action; and acquiring a characteristic representation related to the second characteristic representation in the first characteristic representation as a third characteristic representation, and predicting a first prediction result of the operation behavior of the user on the article. According to the method and the device, the correlation between the upstream behavior and the downstream behavior is modeled explicitly by solving the correlation characteristics between the upstream behavior and the downstream behavior, so that the migration of harmful information of independent parts in upstream information to the downstream is avoided, and the accuracy of a model is improved.

Description

Data processing method and related device

Technical Field

The present disclosure relates to the field of artificial intelligence, and in particular, to a data processing method and related apparatus.

Background

Artificial intelligence (artificial intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar manner to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.

The recommendation system is widely applied to various online services such as electronic commerce, application stores and the like, and aims to provide better services for users and provide maximum benefits for platforms. Researchers have proposed advanced recommendation algorithms based on historical behavior of users. In an actual scenario, there are a variety of user behaviors, such as users interacting with items by browsing, joining shopping carts, purchasing, etc. Therefore, recent work has considered various behaviors in the recommendation model in order to obtain better performance. Among the various activities of users, there is typically one particular activity (e.g., purchase) that we are most interested in, which has a significant and direct relationship with the campaigns of the business platform and with the experience of the improving user, defined as the target activity. In such cases, to better predict the target behavior of the user, researchers typically explore other auxiliary behaviors of the user (e.g., browse, join shopping carts, etc.) to help predict the target behavior.

Early multi-behavioral studies fully utilized auxiliary behavior by expanding matrix decomposition or adjusting sampling strategies in order to fully utilize various behaviors. However, these approaches cannot effectively capture the complex dependencies of multiple behaviors due to the simple architecture. Alternatively, although such hierarchical progression between multiple behaviors is considered, the correlation between different behaviors is modeled "implicitly" by information transfer only through task-specific extractors, resulting in a negative migration phenomenon, which results in lower recommendation accuracy of the recommendation model.

Disclosure of Invention

The data processing method can improve the recommendation precision of the trained recommendation model.

In a first aspect, the present application provides a data processing method, the method comprising: acquiring a first feature representation and a second feature representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is an associated action on the article before the user performs the second operation action; acquiring a feature representation related to the second feature representation in the first feature representation as a third feature representation; predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second characteristic representation and the third characteristic representation; and updating the first characteristic representation and the second characteristic representation according to the first prediction result.

In the embodiment of the application, in order to alleviate negative transfer, the correlation between the upstream and downstream behaviors is modeled explicitly by using a projection mechanism by solving the correlation characteristics between the upstream behavior and the downstream behavior (namely, extracting the shared information in the characterization of the upstream behavior and the downstream behavior from the upstream behavior characterization), and only the thinned shared characterization is transferred, so that the migration of harmful information of an independent part in the upstream information to the downstream is avoided, and the recommendation precision of a trained recommendation model can be improved.

Wherein the attribute information of the user may be at least one of attribute related to preference characteristics of the user, gender, age, occupation, income, hobbies, and education level, wherein the gender may be male or female, the age may be a number between 0 and 100, the occupation may be teacher, programmer, chef, etc., the hobbies may be basketball, tennis, running, etc., and the education level may be primary school, junior middle school, high school, university, etc.; the present application is not limited to a specific type of attribute information of the user.

The article may be a physical article or a virtual article, for example, may be an article such as APP, audio/video, web page, news information, etc., and the attribute information of the article may be at least one of an article name, a developer, an installation package size, a class, and a good score, where, taking the article as an application, the class of the article may be a chat class, a running game, an office class, etc., and the good score may be a score, a comment, etc. for the article; the present application is not limited to a particular type of attribute information for an item.

The operation type may be a behavior operation type of the user aiming at the article, and on the network platform and application, the user often has various interaction forms (i.e. multiple operation types) with the article, such as browsing, clicking, adding shopping cart, purchasing and other operation types of the user in the behavior of the e-commerce platform. These diverse behaviors reflect the user's preferences and are helpful in accurately characterizing the user.

In one possible implementation, the first operation behavior is an action performed on the article before the user performs the second operation behavior, and the first operation behavior is an associated action of the second operation behavior, that is, the first operation behavior is an operation behavior that is performed with a high probability or necessarily needs to be performed before the second operation behavior is performed, for example, the user needs to browse the article before joining in a shopping cart or collecting, or does not make a purchase after joining in the shopping cart or collecting, or does not make a purchase until the user needs to browse the article after joining in the shopping cart or collecting.

In one possible implementation, the obtaining the feature representation related to the second feature representation in the first feature representation is a third feature representation, including: and projecting the first characteristic representation to the direction in which the second characteristic representation is located, so as to obtain a third characteristic representation.

In one possible implementation, the first operational behavior is a browse operation and the second operational behavior is to join a shopping cart, collect or purchase; or the first operation behavior is to join a shopping cart or collect, and the second operation behavior is to purchase; alternatively, the first operation behavior is clicking or browsing, and the second operation behavior is downloading or using.

In one possible implementation, the method further comprises: predicting a second prediction result of the operation behavior of the user on the article according to the second characteristic representation; said updating said first and second feature representations according to said first prediction result, comprising: and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the second prediction result.

In one possible implementation, the method further comprises: predicting a third prediction result of the operation behavior of the user on the article according to the first characteristic representation; said updating said first and second feature representations according to said first prediction result, comprising: and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the third prediction result.

In one possible implementation, the method further comprises: acquiring a feature representation which is irrelevant to the second feature representation in the first feature representation as a fourth feature representation; acquiring a feature representation related to the first feature representation in the fourth feature representation as a fifth feature representation; predicting a fourth prediction result of the operation behavior of the user on the article according to the fusion result of the fifth feature representation and the first feature representation; said updating said first and second feature representations according to said first prediction result, comprising: and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the fourth prediction result.

In one possible implementation, the obtaining a feature representation of the first feature representation that is not related to the second feature representation is a fourth feature representation, including: and projecting the first characteristic representation to the orthogonal direction of the direction in which the second characteristic representation is positioned, so as to obtain a fourth characteristic representation.

The embodiment of the application provides a reprojection method, which utilizes reprojection to transmit independent characterization to the upstream, and then transmits the generated reprojection characterization to the upstream through an extraction network, so that independent interaction information of the upstream is focused more from the aspect of feature enhancement, and the mixing of the independent information of the upstream into shared information transmitted from the downstream is avoided.

In one possible implementation, the method further comprises: predicting a fifth prediction result of the second operation behavior of the article by the user in the presence of the first operation behavior or predicting a fifth prediction result of the first operation behavior of the article by the user in the presence of the second operation behavior according to the fourth characteristic representation; said updating said first and second feature representations according to said first prediction result, comprising: and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the fifth prediction result.

In this way, tag information of the independent portion is utilized, while learning of the complementary shared portion is facilitated.

In one possible implementation, the method further comprises: acquiring a sixth feature representation; the sixth feature is expressed as being obtained based on attribute information of the user and the article and third operation information of the user on the article, wherein the third operation information indicates whether the third operation behavior exists on the article or not, and the second operation behavior is an associated behavior on the article before the third operation behavior is carried out by the user; acquiring a feature representation which is irrelevant to the sixth feature representation in the second feature representation as a seventh feature representation; acquiring a feature representation related to the second feature representation in the seventh feature representation as an eighth feature representation; the predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second feature representation and the third feature representation comprises: and predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second feature representation, the third feature representation and the eighth feature representation.

That is, the feature representation corresponding to the upstream behavior and the feature representation corresponding to the midstream behavior may be fused among the feature shared by the feature representation corresponding to the midstream behavior, and the reprojection feature of the downstream behavior to midstream behavior (the upstream, midstream, and downstream are relative to each other, for example, the upstream behavior is the first operation behavior, the midstream behavior is the second operation behavior, and the downstream behavior is the third operation behavior).

In one possible implementation, the method further comprises: predicting a sixth prediction result of the user's presence of an operational behavior on the item according to the sixth feature representation; said updating said first and second feature representations according to said first prediction result, comprising:

and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the sixth prediction result.

In a second aspect, the present application provides a data processing apparatus, the apparatus comprising:

the processing module is used for acquiring a first characteristic representation and a second characteristic representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is an associated action on the article before the user performs the second operation action;

Acquiring a feature representation related to the second feature representation in the first feature representation as a third feature representation;

predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second characteristic representation and the third characteristic representation;

and the updating module is used for updating the first characteristic representation and the second characteristic representation according to the first prediction result.

In one possible implementation, the processing module is specifically configured to:

and projecting the first characteristic representation to the direction in which the second characteristic representation is located, so as to obtain a third characteristic representation.

In one possible implementation of the present invention,

the first operation behavior is a browsing operation, and the second operation behavior is the addition of a shopping cart, collection or purchase; or alternatively, the process may be performed,

the first operation behavior is to join a shopping cart or collect, and the second operation behavior is to purchase; or alternatively, the process may be performed,

the first operation behavior is clicking or browsing, and the second operation behavior is downloading or using.

In one possible implementation, the processing module is further configured to:

predicting a second predicted result of the user's presence of an operational behavior on the item according to the second characteristic representation;

The updating module is specifically configured to:

and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the second prediction result.

In one possible implementation, the processing module is further configured to:

predicting a third prediction result of the operation behavior of the user on the article according to the first characteristic representation;

the updating module is specifically configured to:

and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the third prediction result.

In one possible implementation, the processing module is further configured to:

acquiring a feature representation which is irrelevant to the second feature representation in the first feature representation as a fourth feature representation; acquiring a feature representation related to the first feature representation in the fourth feature representation as a fifth feature representation;

predicting a fourth prediction result of the operation behavior of the user on the article according to the fusion result of the fifth feature representation and the first feature representation;

the updating module is specifically configured to:

and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the fourth prediction result.

and projecting the first characteristic representation to the orthogonal direction of the direction in which the second characteristic representation is positioned, so as to obtain a fourth characteristic representation.

In one possible implementation, the processing module is further configured to:

predicting a fifth prediction result of the second operation behavior of the article by the user in the presence of the first operation behavior or predicting a fifth prediction result of the first operation behavior of the article by the user in the presence of the second operation behavior according to the fourth characteristic representation;

the updating module is specifically configured to:

and updating the first characteristic representation and the second characteristic representation according to the first prediction result and the fifth prediction result.

In one possible implementation, the processing module is further configured to:

acquiring a sixth feature representation; the sixth feature is expressed as being obtained based on attribute information of the user and the article and third operation information of the user on the article, wherein the third operation information indicates whether the third operation behavior exists on the article or not, and the second operation behavior is an associated behavior on the article before the third operation behavior is carried out by the user;

Acquiring a feature representation which is irrelevant to the sixth feature representation in the second feature representation as a seventh feature representation; acquiring a feature representation related to the second feature representation in the seventh feature representation as an eighth feature representation;

the predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second feature representation and the third feature representation comprises:

and predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second feature representation, the third feature representation and the eighth feature representation.

In one possible implementation, the processing module is further configured to:

predicting a sixth prediction result of the user's presence of an operational behavior on the item according to the sixth feature representation;

the updating module is specifically configured to:

In a third aspect, an embodiment of the present application provides a data processing method, including:

and obtaining the recommendation information of the user for the article through a recommendation model by utilizing the updated characteristic representation of the user and the updated characteristic representation of the article obtained in the first aspect, and determining to recommend the article to the user when the recommendation information meets the preset condition.

And in actual reasoning, the updated characteristic representation of the user and the updated characteristic representation of the article can be utilized to obtain the recommendation information of the user on the article through a recommendation model, and when the recommendation information meets the preset condition, the article can be determined to be recommended to the user.

The preset conditions are described next:

in one possible implementation, when information recommendation is performed on a user, a probability that the user performs multiple operation types on multiple items (including the items) can be calculated, and a recommendation index of each item for the user can be determined based on the probabilities of the multiple operation types.

In one possible implementation, a maximum probability of a probability of multiple operation types of the respective item by the user may be selected to characterize a recommendation index of the respective item to the user;

in one possible implementation, a comprehensive value of probabilities of multiple operation types of the user on each article may be calculated to characterize a recommendation index of each article on the user, where the comprehensive value may be based on a weighted summation mode, specifically, a corresponding weight may be set for each operation type, for example, a weight of a purchase operation is greater than a weight of an operation of joining a shopping cart, and then the recommendation index of each operation type may be obtained based on the weighted summation by combining the weight corresponding to each operation type and the probability corresponding to each operation type;

After the recommendation index for the user for each item is obtained, the recommendation indexes may be ranked and the M items (including the items) with the largest recommendation index may be recommended to the user.

In one possible implementation, a probability threshold may be optionally set, and when a probability that at least one of multiple operation types of the article corresponds to a probability that is greater than the probability threshold, the article may be recommended to the user.

When information recommendation is performed, recommendation information can be recommended to a user in the form of list pages so as to expect the user to perform behavior actions.

In a fourth aspect, embodiments of the present application provide a data processing apparatus, which may include a memory, a processor, and a bus system, where the memory is configured to store a program, and the processor is configured to execute the program in the memory, so as to perform any of the optional methods according to the first aspect.

In a fifth aspect, embodiments of the present application provide a data processing apparatus, which may include a memory, a processor, and a bus system, where the memory is configured to store a program, and the processor is configured to execute the program in the memory, so as to perform any of the optional methods according to the second aspect.

In a sixth aspect, embodiments of the present application provide a computer readable storage medium having a computer program stored therein, which when run on a computer, causes the computer to perform the first aspect and any of the optional methods described above, and any of the optional methods described in the second aspect described above.

In a seventh aspect, embodiments of the present application provide a computer program product comprising code which, when executed, is adapted to carry out the first aspect and any optional method described above, and any optional method of the second aspect described above.

In an eighth aspect, the present application provides a chip system comprising a processor for supporting an execution device or training device to perform the functions involved in the above aspects, e.g. to send or process data involved in the above method; or, information. In one possible design, the chip system further includes a memory for holding program instructions and data necessary for the execution device or the training device. The chip system can be composed of chips, and can also comprise chips and other discrete devices.

Drawings

FIG. 1 is a schematic diagram of a structure of an artificial intelligence main body frame;

FIG. 2 is a schematic diagram of a system architecture according to an embodiment of the present disclosure;

fig. 3 is a schematic diagram of an information recommendation flow provided in an embodiment of the present application;

fig. 4 is a flow chart of an information data processing method according to an embodiment of the present application;

fig. 5 is a schematic diagram of a network architecture according to an embodiment of the present application;

fig. 6 is a schematic view of an application scenario provided in an embodiment of the present application;

FIG. 7 is a schematic diagram of a data processing apparatus according to an embodiment of the present application;

fig. 8 is a schematic diagram of an execution device according to an embodiment of the present application;

FIG. 9 is a schematic diagram of a training device according to an embodiment of the present application;

fig. 10 is a schematic diagram of a chip according to an embodiment of the present application.

Detailed Description

Embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention. The terminology used in the description of the embodiments of the invention herein is for the purpose of describing particular embodiments of the invention only and is not intended to be limiting of the invention.

Embodiments of the present application are described below with reference to the accompanying drawings. As one of ordinary skill in the art can appreciate, with the development of technology and the appearance of new scenes, the technical solutions provided in the embodiments of the present application are applicable to similar technical problems.

The terms first, second and the like in the description and in the claims of the present application and in the above-described figures, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and are merely illustrative of the manner in which the embodiments of the application described herein have been described for objects of the same nature. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of elements is not necessarily limited to those elements, but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.

Referring to fig. 1, a schematic structural diagram of an artificial intelligence main body framework is shown in fig. 1, and the artificial intelligence main body framework is described below from two dimensions of "intelligent information chain" (horizontal axis) and "IT value chain" (vertical axis). Where the "intelligent information chain" reflects a list of processes from the acquisition of data to the processing. For example, there may be general procedures of intelligent information awareness, intelligent information representation and formation, intelligent reasoning, intelligent decision making, intelligent execution and output. In this process, the data undergoes a "data-information-knowledge-wisdom" gel process. The "IT value chain" reflects the value that artificial intelligence brings to the information technology industry from the underlying infrastructure of personal intelligence, information (provisioning and processing technology implementation), to the industrial ecological process of the system.

(1) Infrastructure of

The infrastructure provides computing capability support for the artificial intelligence system, realizes communication with the outside world, and realizes support through the base platform. Communicating with the outside through the sensor; the computing power is provided by a smart chip (CPU, NPU, GPU, ASIC, FPGA and other hardware acceleration chips); the basic platform comprises a distributed computing framework, a network and other relevant platform guarantees and supports, and can comprise cloud storage, computing, interconnection and interworking networks and the like. For example, the sensor and external communication obtains data that is provided to a smart chip in a distributed computing system provided by the base platform for computation.

(2) Data

The data of the upper layer of the infrastructure is used to represent the data source in the field of artificial intelligence. The data relate to graphics, images, voice and text, and also relate to the internet of things data of the traditional equipment, including service data of the existing system and sensing data such as force, displacement, liquid level, temperature, humidity and the like.

(3) Data processing

Data processing typically includes data training, machine learning, deep learning, searching, reasoning, decision making, and the like.

Wherein machine learning and deep learning can perform symbolized and formalized intelligent information modeling, extraction, preprocessing, training and the like on data.

Reasoning refers to the process of simulating human intelligent reasoning modes in a computer or an intelligent system, and carrying out machine thinking and problem solving by using formal information according to a reasoning control strategy, and typical functions are searching and matching.

Decision making refers to the process of making decisions after intelligent information is inferred, and generally provides functions of classification, sequencing, prediction and the like.

(4) General capability

After the data has been processed, some general-purpose capabilities can be formed based on the result of the data processing, such as algorithms or a general-purpose system, for example, translation, text analysis, computer vision processing, speech recognition, image recognition, etc.

(5) Intelligent product and industry application

The intelligent product and industry application refers to products and applications of an artificial intelligent system in various fields, is encapsulation of an artificial intelligent overall solution, and realizes land application by making intelligent information decisions, and the application fields mainly comprise: intelligent terminal, intelligent transportation, intelligent medical treatment, autopilot, smart city etc.

The embodiment of the application can be applied to the field of information recommendation, wherein the scenes comprise but are not limited to scenes related to e-commerce product recommendation, search engine result recommendation, application market recommendation, music recommendation, video recommendation and the like, the recommended objects in various application scenes can be also called as 'objects' so as to facilitate subsequent description, namely in different recommendation scenes, the recommended objects can be APP, video, or music, or a certain commodity (such as a presentation interface of an online shopping platform, different commodities can be displayed according to different users to be presented), and the essence can also be presented through the recommendation result of a recommendation model. These recommendation scenarios typically involve user behavior log collection, log data preprocessing (e.g. quantization, sampling, etc.), sample set training to obtain recommendation models, analysis of the objects (e.g. APP, music, etc.) involved in the scenario to which the training sample items correspond according to the recommendation models, e.g. the samples selected in the recommendation model training session come from the mobile phone APP application market user's operation behavior on the recommended APP, the recommendation models thus trained are then applicable to the mobile phone APP application market described above, or the APP application market for other types of terminals may be used to make recommendations of the terminal APP. The recommendation model finally calculates the recommendation probability or score of each object to be recommended, the recommendation system sorts the recommendation results selected according to a certain selection rule, for example, the recommendation results are sorted according to the recommendation probability or score, and the recommendation results are presented to the user through corresponding application or terminal equipment, and the user operates the objects in the recommendation results to generate links such as user behavior logs.

In the recommendation process, when a user interacts with the recommendation system, a recommendation request is triggered, the recommendation system inputs the request and related characteristic information into a deployed recommendation model, and then the click rate of the user on all candidate objects is predicted. And then, the candidate objects are arranged in a descending order according to the predicted click rate, and the candidate objects are displayed at different positions in order to serve as recommendation results for users. The user browses the presented items and user behavior such as browsing, clicking, downloading, etc. occurs. The user behaviors can be stored in a log to be used as training data, and the parameters of the recommendation model are updated irregularly through the offline training module, so that the recommendation effect of the model is improved.

For example, a user opens a mobile phone application market to trigger a recommendation module of the application market, and the recommendation module of the application market predicts the possibility of downloading given candidate applications by the user according to the historical downloading records of the user, the clicking records of the user, the self-characteristics of the applications, the time, the place and other environmental characteristic information. According to the predicted result, the application market is displayed according to the descending order of the possibility, and the effect of improving the application downloading probability is achieved. Specifically, applications that are more likely to be downloaded are ranked in a front position, and applications that are less likely to be downloaded are ranked in a rear position. The behavior of the user is also logged and the parameters of the prediction model are trained and updated through the offline training module.

For example, in the application related to life mate, the cognitive brain can be built by simulating the brain mechanism through various models and algorithms based on the historical data of the user in the fields of video, music, news and the like, and the life learning system framework of the user is built. The life mate can record events occurring in the past of the user according to system data, application data and the like, understand the current intention of the user, predict future actions or behaviors of the user and finally realize intelligent service. In the current first stage, behavior data (including information such as terminal side short messages, photos and mail events) of a user are obtained according to a music APP, a video APP, a browser APP and the like, a user portrait system is built, and learning and memory modules based on user information filtering, association analysis, cross-domain recommendation, causal reasoning and the like are realized to build a user personal knowledge map.

Next, an application architecture of the embodiment of the present application is described.

Referring to fig. 2, an embodiment of the present invention provides a recommendation system architecture 500. The data collection device 560 is configured to collect samples, where one training sample may be composed of a plurality of feature information (or be described as attribute information, such as user attribute and article attribute), and the feature information may include user feature information and object feature information, and tag feature, where the user feature information is used to characterize a feature of a user, such as gender, age, occupation, hobbies, etc., the object feature information is used to characterize a feature of an object pushed to the user, different recommendation systems correspond to different objects, and types of features that need to be extracted by different objects are also not wanted, for example, the object feature extracted in the training sample of the APP market may be a name (identifier), a type, a size, etc. of APP; the object features mentioned in the training sample of the e-commerce APP can be the names of commodities, the category to which the commodities belong, price intervals and the like; the label feature is used to indicate whether the sample is a positive example or a negative example, and in general, the label feature of the sample may be obtained through operation information of the recommended object by the user, the sample in which the user has operated the recommended object is a positive example, the recommended object is not operated by the user, or only the sample browsed is a negative example, for example, when the user clicks or downloads or purchases the recommended object, the label feature is 1, which indicates that the sample is a positive example, and if the user has not operated any recommended object, the label feature is 0, which indicates that the sample is a negative example. In this application, the training sample may also include the type of user manipulation of the item. The samples may be stored in the database 530 after collection, and some or all of the characteristic information in the samples in the database 530 may also be obtained directly from the client device 540, such as user characteristic information, user operation information on the object (for determining a type identifier), object characteristic information (such as an object identifier), and so on. The training device 520 trains the acquisition model parameter matrix based on the samples in the database 530 for generating the recommendation model 501. How the training device 520 trains to obtain the model parameter matrix for generating the recommendation model 501 is described in more detail below, the recommendation model 501 can be used to evaluate a large number of objects to obtain the score of each object to be recommended, further a specified or preset number of objects can be recommended from the evaluation results of the large number of objects, and the computing module 511 obtains the recommendation result based on the evaluation results of the recommendation model 501 and recommends the recommendation result to the client device through the I/O interface 512.

In this embodiment, the training device 520 may select positive samples and negative samples from the sample set in the database 530 and add the positive samples and the negative samples to the training set, and then train the samples in the training set by using the recommendation model to obtain a trained recommendation model; details of the implementation of the computing module 511 may be found in the detailed description of the method embodiment shown in fig. 5.

The training device 520 is used for constructing the recommendation model 501 after obtaining the model parameter matrix based on sample training, and then sending the recommendation model 501 to the execution device 510, or directly sending the model parameter matrix to the execution device 510, and constructing a recommendation model in the execution device 510 for recommending a corresponding system, for example, the recommendation model obtained based on sample training related to video can be used for recommending video to a user in a video website or an APP, and the recommendation model obtained based on sample training related to APP can be used for recommending APP to the user in an application market.

The execution device 510 is configured with an I/O interface 512, and performs data interaction with an external device, and the execution device 510 may obtain user characteristic information, such as a user identifier, a user identity, a gender, a occupation, a preference, etc., from the client device 540 through the I/O interface 512, and this part of information may also be obtained from a system database. The recommendation model 501 recommends a target recommended object to the user based on the user characteristic information and the object characteristic information to be recommended. The execution device 510 may be disposed in a cloud server or in a user client.

The execution device 510 may invoke data, code, etc. in the data storage system 550 and may store the output data in the data storage system 550. The data storage system 550 may be disposed in the execution device 510, may be disposed independently, or may be disposed in other network entities, and the number may be one or multiple.

The computing module 511 processes the user feature information by using the recommendation model 501, and for example, the computing module 511 analyzes the user feature information and the feature information of the object to be recommended by using the recommendation model 501, so as to obtain the score of the object to be recommended, and sorts the objects to be recommended according to the score, wherein the object with the front sorting is to be the object recommended to the client device 540.

Finally, the I/O interface 512 returns the recommendation to the client device 540 for presentation to the user.

Further, the training device 520 may generate respective recommendation models 501 based on different sample feature information for different targets to provide better results to the user.

It should be noted that fig. 2 is only a schematic diagram of a system architecture according to an embodiment of the present invention, and the positional relationship among devices, apparatuses, modules, etc. shown in the drawing is not limited in any way, for example, in fig. 2, the data storage system 550 is an external memory with respect to the execution device 510, and in other cases, the data storage system 550 may be disposed in the execution device 510.

In this embodiment of the present application, the training device 520, the executing device 510, and the client device 540 may be three different physical devices, or the training device 520 and the executing device 510 may be on the same physical device or a cluster, or the executing device 510 and the client device 540 may be on the same physical device or a cluster.

The execution device 510 may be implemented by one or more servers, optionally in conjunction with other computing devices, such as: data storage, routers, load balancers and other devices; the execution device 510 may be disposed on one physical site or distributed across multiple physical sites. The executing device 510 may use data in the data storage system 550 or call program codes in the data storage system 550 to implement an object recommendation function, specifically, input information of an object to be recommended into a recommendation model, generate a pre-estimated score for each object to be recommended by the recommendation model, sort the objects according to the pre-estimated score from high to low, and recommend the object to be recommended to a user according to the sorting result. For example, the first 10 objects in the ranking result are recommended to the user.

The data storage system 550 is configured to receive and store parameters of the recommendation model sent by the training device, and data for storing recommendation results obtained by the recommendation model, and may also include program code (or instructions) required for the normal operation of the storage system 550. The data storage system 550 may be a distributed storage cluster formed by one device or a plurality of devices disposed outside the execution device 510, where when the execution device 510 needs to use the data on the storage system 550, the data required by the execution device may be sent by the storage system 550 to the execution device 510, and accordingly, the execution device 510 receives and stores (or caches) the data. Of course, the data storage system 550 may also be deployed within the execution device 510, and when deployed within the execution device 510, the distributed storage system may include one or more memories, and optionally, where there are multiple memories, different memories may be used to store different types of data, such as model parameters of a recommendation model generated by the training device and data of recommendation results obtained by the recommendation model, may be stored on two different memories, respectively.

The user may operate a respective user device (e.g., client device 540) to interact with the execution device 510. Each local device may represent any computing device, such as a personal computer, computer workstation, smart phone, tablet, smart camera, smart car or other type of cellular phone, media consumption device, wearable device, set top box, game console, etc.

The local device of each user may interact with the performing device 510 via a communication network of any communication mechanism/communication standard, which may be a wide area network, a local area network, a point-to-point connection, etc., or any combination thereof.

In another implementation, the execution device 510 may be implemented by a local device.

Since the embodiments of the present application relate to a large number of applications of neural networks, for ease of understanding, related terms and related concepts of the neural networks related to the embodiments of the present application will be described below.

1. Click-through probability (CTR)

The click probability, which may also be referred to as a click rate, refers to the ratio of the number of clicks to the number of exposures of recommended information (e.g., recommended items) on a web site or application, and is typically an important indicator in a recommendation system to measure the recommendation system.

2. Personalized recommendation system

The personalized recommendation system is a system for analyzing according to historical data (such as operation information in the embodiment of the application) of a user by using a machine learning algorithm, predicting a new request according to the analysis, and giving a personalized recommendation result.

3. Offline training (offlinenet)

The offline training refers to a module for iteratively updating recommendation model parameters according to an algorithm learned by a device according to historical data (such as operation information in the embodiment of the application) of a user in a personalized recommendation system until the recommendation model parameters reach a set requirement.

4. Online prediction (onlineiference)

The online prediction refers to predicting the preference degree of the user for the recommended item in the current context according to the characteristics of the user, the item and the context based on an offline trained model, and predicting the probability of selecting the recommended item by the user.

5. Multi-behavior recommendation system: and analyzing and modeling by utilizing a machine learning algorithm according to the multi-behavior historical interaction behavior data of the user, predicting a new user request according to the analysis and modeling, and giving a personalized recommendation result.

6. Negative migration phenomenon: the negative migration phenomenon represents a phenomenon that the performance of a model is reduced when harmful information is transmitted between different tasks. Information affecting learning of the target behavior is defined as harmful information.

7. Multitasking learning: the multi-task learning is a paradigm for processing different tasks by using the same characterization, and the combined optimization mode is widely used in the fields of recommendation and the like, so that a model can learn complex multi-source heterogeneous information better.

For example, fig. 3 is a schematic diagram of a recommendation system provided in an embodiment of the present application. As shown in FIG. 3, when a user enters the system, a request for a recommendation is triggered, the recommendation system enters the request and its associated information into the recommendation model, and then predicts the user's selectivity (or probability of performing a particular operational action) for items within the system. Further, the items are arranged in a descending order according to the predicted selectivity or some function based on the selectivity, i.e. the recommendation system may display the items in different positions in order as a recommendation to the user. The user browses the different items in place and user actions occur such as browsing, selecting, downloading, etc. Meanwhile, the actual behaviors of the user can be stored in a log to be used as training data, and parameters of the recommendation model are continuously updated through the offline training module, so that the prediction effect of the model is improved.

For example, a user opening an application marketplace in a smart terminal (e.g., a cell phone) may trigger a recommendation system in the application marketplace. The recommendation system of the application market predicts the probability of downloading recommended candidate APP by the user according to the historical behavior log of the user, for example, the historical downloading record of the user, the user selection record and the self-characteristics of the application market, such as time, place and other environmental characteristic information. According to the calculated result, the recommendation system of the application market can display the candidate APP in descending order according to the predicted probability value, so that the downloading probability of the candidate APP is improved.

For example, APP with higher predicted user selectivity may be shown at a forward recommended position and APP with lower predicted user selectivity may be shown at a rearward recommended position.

The recommendation model may be a neural network model, and related terms and concepts of the neural network that may be related to the embodiments of the present application are described below.

(1) Neural network

The neural network may be composed of neural units, which may refer to an arithmetic unit with xs (i.e., input data) and intercept 1 as inputs, and the output of the arithmetic unit may be:

where s=1, 2, … … n, n is a natural number greater than 1, ws is the weight of xs, and b is the bias of the neural unit. f is an activation function (activation functions) of the neural unit for introducing a nonlinear characteristic into the neural network to convert an input signal in the neural unit to an output signal. The output signal of the activation function may be used as an input to a next convolutional layer, and the activation function may be a sigmoid function. A neural network is a network formed by joining together a plurality of the above-described single neural units, i.e., the output of one neural unit may be the input of another neural unit. The input of each neural unit may be connected to a local receptive field of a previous layer to extract features of the local receptive field, which may be an area composed of several neural units.

(2) Deep neural network

Deep neural networks (Deep Neural Network, DNN), also known as multi-layer neural networks, can be understood as neural networks having many hidden layers, many of which are not particularly metrics. From DNNs, which are divided by the location of the different layers, the neural networks inside the DNNs can be divided into three categories: input layer, hidden layer, output layer. Typically the first layer is the input layer, the last layer is the output layer, and the intermediate layers are all hidden layers. The layers are fully connected, that is, any neuron in the i-th layer must be connected to any neuron in the i+1-th layer. Although DNN appears to be complex, it is not really complex in terms of the work of each layer, simply the following linear relational expression:wherein (1)>Is an input vector, +.>Is the output vector, +.>Is the offset vector, W is the weight matrix (also called coefficient), and α () is the activation function. Each layer is only for the input vector +.>The output vector is obtained by such simple operation>Since DNN has a large number of layers, the coefficient W and the offset vector +.>And thus a large number. The definition of these parameters in DNN is as follows: taking the coefficient W as an example: it is assumed that in DNN of one three layers, the linear coefficients of the 4 th neuron of the second layer to the 2 nd neuron of the third layer are defined as +. >The superscript 3 represents the number of layers in which the coefficient W is located, and the subscript corresponds to the output third layer index 2 and the input second layer index 4. The summary is: the coefficients from the kth neuron of the L-1 th layer to the jth neuron of the L-1 th layer are defined as +.>It should be noted that the input layer is devoid of W parameters. In deep neural networks, more hidden layers make the network more capable of characterizing complex situations in the real world. Theoretically, the more parameters the higher the model complexity, the greater the "capacity", meaning that it can be doneA more complex learning task. The process of training the deep neural network, i.e. learning the weight matrix, has the final objective of obtaining a weight matrix (a weight matrix formed by a number of layers of vectors W) for all layers of the trained deep neural network.

(3) Loss function

In training the deep neural network, since the output of the deep neural network is expected to be as close to the value actually expected, the weight vector of each layer of the neural network can be updated by comparing the predicted value of the current network with the actually expected target value according to the difference between the predicted value of the current network and the actually expected target value (of course, there is usually an initialization process before the first update, that is, the pre-configuration parameters of each layer in the deep neural network), for example, if the predicted value of the network is higher, the weight vector is adjusted to be predicted to be lower, and the adjustment is continued until the deep neural network can predict the actually expected target value or the value very close to the actually expected target value. Thus, it is necessary to define in advance "how to compare the difference between the predicted value and the target value", which is a loss function (loss function) or an objective function (objective function), which are important equations for measuring the difference between the predicted value and the target value. Taking the loss function as an example, the higher the output value (loss) of the loss function is, the larger the difference is, and then the training of the deep neural network becomes a process of reducing the loss as much as possible.

(4) Back propagation algorithm

An error Back Propagation (BP) algorithm may be used to correct the magnitude of the parameters in the initial model during the training process, so that the error loss of the model is smaller and smaller. Specifically, the input signal is forward-transferred until output, and error loss occurs, and parameters in the initial model are updated by back-propagating the error loss information, so that the error loss converges. The back propagation algorithm is a back propagation motion that dominates the error loss, aiming at deriving optimal model parameters, such as a weight matrix.

(5) Graph (Graph):

the diagram is a data structure including at least one vertex and at least one edge. In some scenarios, vertices in the graph may map as entities, and edges in the graph may map as entities to relationships between entities. The graph may be a directed graph or an undirected graph. Of course, the graph may also include other data besides vertices and edges, such as labels for vertices and labels for edges. In one exemplary scenario, applied to the recommended scenario, each vertex in the graph may represent a user or an item, each edge in the graph may represent a user interaction relationship with the item, such as a purchase relationship, a collection relationship, a click relationship, etc., and data of each vertex in the graph is attribute information of the user, such as age, occupation, hobbies, academic, etc., of the user.

(6) Fig. neural network (graph neural network, GNN):

GNN is a deep learning method with structural information that can be used to calculate the current state of a node. The information transmission of the graph neural network is carried out according to a given graph structure, and the state of each node can be updated according to adjacent nodes. Specifically, according to the structure diagram of the current node, the information of all adjacent nodes can be transferred to the current node by taking the neural network as an aggregation function of point information, and the information is updated by combining the state of the current node. The output of the graph neural network is the state of all nodes.

The data processing method provided in the embodiment of the present application will be described below by taking a model reasoning stage as an example.

Referring to fig. 4, fig. 4 is an embodiment schematic diagram of a data processing method provided in an embodiment of the present application, and as shown in fig. 4, the data processing method provided in the embodiment of the present application includes:

401. acquiring a first feature representation and a second feature representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is the association action performed on the article before the user performs the second operation action.

In an embodiment of the present application, the execution subject of step 401 may be a terminal device, which may be a portable mobile device, such as, but not limited to, a mobile or portable computing device (e.g., a smart phone), a personal computer, a server computer, a handheld device (e.g., a tablet) or laptop, a multiprocessor system, a game console or controller, a microprocessor-based system, a set top box, a programmable consumer electronics, a mobile phone, a mobile computing and/or communication device with a wearable or accessory form factor (e.g., a watch, glasses, a headset, or an earplug), a network PC, a minicomputer, a mainframe computer, a distributed computing environment that includes any of the above systems or devices, and the like.

In this embodiment, the execution body of step 401 may be a cloud server.

For convenience of description, the form of the execution subject is not distinguished, and is described as an execution apparatus. In this embodiment of the present application, the executing device may obtain an operation information set of a plurality of users, where the operation information set may be obtained based on an interaction record (for example, a behavior log of a user) between the user and an article, and information in the operation information set may include a plurality of users and a real operation record of at least one article by each user, and the operation information set may include attribute information of the user, attribute information of each article, and an operation type of an operation behavior (for example, a first operation behavior and a second operation behavior in the embodiment of the present application) performed by the user on the article.

In one possible implementation, feature extraction may be performed on the user, the item, and information indicating that the user has an operation behavior on the item, respectively, to obtain feature representations of the user and the item corresponding to the operation behavior. The characteristic representation of the user and the article corresponding to each operation behavior can comprise the embedded characteristic of each user and the embedded characteristic of each article.

In one possible implementation, taking the first operation behavior and the second operation behavior as examples, the attribute information of the user and the article and the first operation information of the user on the article may be processed through the feature extraction network to obtain a first feature representation, and the attribute information of the user and the article and the second operation information of the user on the article may be processed through the feature extraction network to obtain a second feature representation. Wherein the first feature representation may include an embedded feature for each user and an embedded feature for each item. Wherein the second feature representation may include an embedded feature for each user and an embedded feature for each item.

In one possible implementation, graph data can be constructed based on a set of operational information, nodes in the graph data can correspond to users or items, and edges between nodes can correspond to user operational behavior on items (e.g., whether there is some operational behavior). For example, a user, an item, and information indicating that the user has one operational action on the item may be constructed as one graph, and a user, an item, and information indicating that the user has another operational action on the item may be constructed as another graph. Specifically, reference may be made to fig. 5, where in fig. 5, a graph is constructed from a user, an item, and information indicating that the user has a browsing action on the item, where in the graph, a connection relationship exists between nodes to indicate that the user has a browsing action on the item, in fig. 5, a graph is constructed from a user, an item, and information indicating that the user has a shopping cart adding action on the item, where in the graph, a connection relationship exists between nodes to indicate that the user has a shopping cart adding action on the item, and in fig. 5, a graph is constructed from a user, an item, and information indicating that the user has a purchasing action on the item, where in the graph, a connection relationship exists between nodes to indicate that the user has a purchasing action on the item.

In one possible implementation, feature extraction may be performed on each graph constructed by using the graph neural network to obtain a feature representation corresponding to each graph, where the feature representation corresponding to each graph may include an embedded feature of each user and an embedded feature of each article.

As shown in FIG. 5, the act of feature extraction for each operational action may be performed by a baseline model, illustratively mapping user and merchandise information (e.g., ID numbers) into a low-dimensional continuous token vector through an Embedding layer. And then, taking the user low-dimensional representation x and the commodity low-dimensional representation y as the input of a model, processing the model by a baseline model to obtain decoupling representations of different behaviors, and inputting the decoupling representations into a projection-based migration network.

The specific flow of the input data processing by the baseline model can be divided into the following steps: and (3) using the weight representation of the initialized user and the item, mapping the high-dimensional sparse user and item interaction data into low-dimensional ebedding through an embedding layer, and outputting the ebedding input code network to obtain representations of different behaviors.

402. Acquiring a feature representation related to the second feature representation in the first feature representation as a third feature representation;

In one possible implementation, the first feature representation may be projected in a direction in which the second feature representation is located to obtain a third feature representation, or the related feature representation may be extracted by a deep learning method, which is not limited in this application.

The first feature representation comprises feature embedding of each user and feature embedding of each article, the second feature representation comprises feature embedding of each user and feature embedding of each article, when projection is performed, the feature embedding of each user in the first feature representation can be projected to the direction of the feature embedding of the corresponding user in the second feature representation, and similarly, when projection is performed, the feature embedding of each article in the first feature representation can be projected to the direction of the feature embedding of the corresponding article in the second feature representation.

403. Predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second characteristic representation and the third characteristic representation;

in one possible implementation, the fusion of the second feature representation and the third feature representation may be a summation operation of the corresponding positions, or the like.

404. And updating the first characteristic representation and the second characteristic representation according to the first prediction result.

In one possible implementation, a feature extraction network (e.g., a graph neural network) for deriving the first and second feature representations may be updated in addition to the first and second feature representations.

In one possible implementation, a second prediction of the user's presence of an operational behavior of the item may also be predicted based on the second characteristic representation; further, the first and second feature representations may be updated based on the first and second predictors.

The second prediction result of the operation behavior of the user on the article may be a second prediction result of the operation behavior of the user on the article.

In a possible implementation, a third prediction result of the operation behavior of the user on the article may be further predicted according to the first feature representation; further, the first and second feature representations may be updated based on the first and third predictors.

The third prediction result of the operation behavior of the user on the article may be a third prediction result of the operation behavior of the user on the article.

In a possible implementation, a feature representation of the first feature representation that is not related to the second feature representation may also be obtained as a fourth feature representation; acquiring a feature representation related to the first feature representation in the fourth feature representation as a fifth feature representation; predicting a fourth prediction result of the operation behavior of the user on the article according to the fusion result of the fifth feature representation and the first feature representation; further, the first and second feature representations may be updated based on the first and fourth predictors.

In one possible implementation, the first feature representation may be projected in a direction orthogonal to the direction in which the second feature representation is located, to obtain a fourth feature representation, or the extraction of the unrelated feature representation may be performed by deep learning, which is not limited in this application.

In a possible implementation, a fifth prediction of the second operation behavior of the item by the user in the presence of the first operation behavior or a fifth prediction of the first operation behavior of the item by the user in the presence of the second operation behavior may be predicted based on the fourth characteristic representation; further, the first and second feature representations may be updated based on the first and fifth predictors.

In one possible implementation, a sixth feature representation may be obtained; the sixth feature is expressed as being obtained based on attribute information of the user and the article and third operation information, the third operation information indicates whether the user has a third operation behavior on the article, the third user has a third operation behavior on the third article, and the second operation behavior is an associated behavior on the article before the user performs the third operation behavior; acquiring a feature representation which is irrelevant to the sixth feature representation in the second feature representation as a seventh feature representation; acquiring a feature representation related to the second feature representation in the seventh feature representation as an eighth feature representation; further, a first prediction result of the operation behavior of the user on the article may be predicted based on a fusion result of the second feature representation, the third feature representation, and the eighth feature representation.

In one possible implementation, a sixth prediction of the user's presence of an operational behavior of the item may be predicted based on the sixth characteristic representation; further, the first and second feature representations may be updated based on the first and sixth predictors.

The existing method does not consider hierarchical progressive association among different types of behaviors, only uses coupled characterization as input, processes prediction tasks of different behaviors, and causes the problem of gradient contradiction in the learning process; or by information transfer through task-specific extractors to "implicitly" model the correlation between different behaviors, resulting in negative migration phenomena that occur during learning. From the perspective of multi-behavioral layering progression, embodiments of the present application design a layered projection enhanced multi-behavioral recommendation (HPMR) framework, as shown in fig. 5, in which clients are modeled to include a baseline model and a projection-based migration network (PTN). The baseline model mainly comprises an optional neural network as a backbone to model higher order behavioral dependencies. The PTN mainly includes three components, independent characterization supervision (URS), shared information migration (SIT), and independent characterization re-projection (URR).

The association of upstream and downstream is explicitly modeled using projection mechanisms, and shared and independent characterizations are extracted from the upstream behavioral characterizations. Then, three parts of independent characterization supervision (URS), shared information migration (SIT) and independent characterization re-projection (URR) can be used for respectively processing shared and independent characterization, upstream interaction information is fully extracted and utilized from the aspects of features and labels, prediction of downstream behaviors is enhanced, and the enhanced characterization is output. And calculating and outputting the prediction scores under different tasks by using the refined characterization.

The embodiment of the application provides a data processing method, which comprises the following steps: acquiring a first feature representation and a second feature representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is an associated action on the article before the user performs the second operation action; acquiring a feature representation related to the second feature representation in the first feature representation as a third feature representation; predicting a first prediction result of the operation behavior of the user on the article according to the fusion result of the second characteristic representation and the third characteristic representation; and updating the first characteristic representation and the second characteristic representation according to the first prediction result. In the embodiment of the application, in order to alleviate negative transfer, the correlation between the upstream and downstream behaviors is modeled explicitly by using a projection mechanism by solving the correlation characteristics between the upstream behavior and the downstream behavior (namely, extracting the shared information in the characterization of the upstream behavior and the downstream behavior from the upstream behavior characterization), and only the thinned shared characterization is transferred, so that the migration of harmful information of an independent part in the upstream information to the downstream is avoided, and the recommendation precision of a trained recommendation model can be improved.

The preset conditions are described next:

The method provided by the embodiment of the application is fully tested on a plurality of user multi-behavior recommendation public data sets and industrial data sets, and the test is set as follows:

performance was evaluated using the Beibei, taobao dataset.

Using industry accepted test metrics (HR, higher better) and (NDCG, higher better), several techniques are compared:

(1) Recommendation model based on single behavior: BPR, NCF, ENMF and LightGCN

(2) Multi-behavior model based on single-task learning: CMF, MC-BPR, MBGCN and MATN

(3) Multitasking learning-based multi-behavior model: NMTR, CML, MBGMN, EHCF and GHCF

Table 1 is a schematic representation of a comparison of recommended performance. Wherein bold indicates best results, underline indicates sub-best results (i.e., best baseline). ". Indicates a statistically significant improvement (i.e., p-value < 0.05) over the optimal baseline.

TABLE 1

Through experiments, the following conclusions can be drawn: in terms of recommendation accuracy, the embodiment of the application obtains the best experimental effect on three indexes of Hit Rate and NDCG, and the obvious effectiveness of the multi-behavior recommendation system provided by the embodiment of the application is proved.

In addition, the examples of the present application were also compared to existing work on industrial datasets, and the results are shown in table 2, table 2 being a performance schematic in CTR prediction industrial scenario:

TABLE 2

It can be seen that the embodiments of the present application provide a significant improvement over real industrial data sets, suitable for landed applications.

The system scene applied by the application is an application scene based on machine learning. The description is given by taking a multi-behavior recommendation scene in the recommendation system as an example. The multi-behavior recommendation scene is a typical scene in machine learning application, and the main structure of the multi-behavior recommendation scene is shown in fig. 6, and comprises a presentation list, a log, an offline training module and an online prediction module.

The basic operation logic of the recommendation system is as follows: the user performs a series of actions, such as browsing, clicking, commenting, downloading, etc., in the front-end presentation list, generates action data, and stores the action data in the log. The recommendation system performs offline model training by using data comprising user behavior logs, generates a prediction model after training convergence, deploys the model in an online service environment, gives a recommendation result based on user request access, commodity characteristics and context information, and then generates feedback on the recommendation result by a user to form user data.

The multi-behavior recommendation model models behavior modes and dynamic interests of users through multi-behavior historical interactions between the users and commodities in an offline training module. The user interest modeling is effective and efficient, the accuracy and individuation of the recommendation model can be improved, and the user experience and income are greatly improved.

Referring to fig. 7, fig. 7 is a data processing apparatus 700 provided in an embodiment of the present application, where the apparatus 700 includes:

a processing module 701, configured to obtain a first feature representation and a second feature representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is an associated action on the article before the user performs the second operation action;

for a specific description of the processing module 701, reference may be made to descriptions of step 401, step 402, and step 403, which are not repeated here.

An updating module 702, configured to update the first feature representation and the second feature representation according to the first prediction result.

For a specific description of the update module 702, reference may be made to the description of step 404, which is not repeated here.

In one possible implementation of the present invention,

In one possible implementation, the processing module is further configured to:

the updating module is specifically configured to:

In one possible implementation, the processing module is further configured to:

the updating module is specifically configured to:

In one possible implementation, the processing module is further configured to:

The updating module is specifically configured to:

In one possible implementation, the processing module is further configured to:

the updating module is specifically configured to:

In one possible implementation, the processing module is further configured to:

the updating module is specifically configured to:

Next, referring to fig. 8, fig. 8 is a schematic structural diagram of an execution device provided in the embodiment of the present application, where the execution device 800 may be specifically represented by a mobile phone, a tablet, a notebook computer, an intelligent wearable device, a server, and the like, which is not limited herein. The execution device 800 may be configured with the data processing apparatus described in the corresponding embodiment of fig. 10, to implement the functions of data processing in the corresponding embodiment of fig. 10. Specifically, the execution apparatus 800 includes: a receiver 801, a transmitter 802, a processor 803, and a memory 804 (where the number of processors 803 in the execution device 800 may be one or more), where the processor 803 may include an application processor 8031 and a communication processor 8032. In some embodiments of the present application, the receiver 801, transmitter 802, processor 803, and memory 804 may be connected by a bus or other means.

Memory 804 may include read only memory and random access memory and provides instructions and data to the processor 803. A portion of the memory 804 may also include non-volatile random access memory (NVRAM). The memory 804 stores a processor and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, where the operating instructions may include various operating instructions for performing various operations.

The processor 803 controls the operation of the execution device. In a specific application, the individual components of the execution device are coupled together by a bus system, which may include, in addition to a data bus, a power bus, a control bus, a status signal bus, etc. For clarity of illustration, however, the various buses are referred to in the figures as bus systems.

The methods disclosed in the embodiments of the present application may be applied to the processor 803 or implemented by the processor 803. The processor 803 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuitry of hardware or instructions in software form in the processor 803. The processor 803 may be a general purpose processor, a digital signal processor (digital signal processing, DSP), a microprocessor or microcontroller, a visual processor (vision processing unit, VPU), a tensor processor (tensor processing unit, TPU), etc. suitable for AI operation, and may further include an application specific integrated circuit (application specific integrated circuit, ASIC), a field-programmable gate array (field-programmable gate array, FPGA), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. The processor 803 may implement or perform the methods, steps, and logic blocks disclosed in the embodiments of the present application. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present application may be embodied directly in hardware, in a decoded processor, or in a combination of hardware and software modules in a decoded processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in the memory 804, and the processor 803 reads the information in the memory 804, and in combination with the hardware thereof, performs the steps related to model reasoning in the above embodiment.

The receiver 801 may be used to receive input numeric or character information and to generate signal inputs related to performing relevant settings and function control of the device. The transmitter 802 may be used to output numeric or character information through a first interface; the transmitter 802 may also be configured to send instructions to the disk group through the first interface to modify data in the disk group; the transmitter 802 may also include a display device such as a display screen.

Referring to fig. 9, fig. 9 is a schematic structural diagram of the training device provided in the embodiment of the present application, specifically, the training device 900 is implemented by one or more servers, where the training device 900 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (central processing units, CPU) 99 (e.g., one or more processors) and a memory 932, and one or more storage media 930 (e.g., one or more mass storage devices) storing application programs 942 or data 944. Wherein the memory 932 and the storage medium 930 may be transitory or persistent. The program stored on the storage medium 930 may include one or more modules (not shown), each of which may include a series of instruction operations for the training device. Still further, the central processor 99 may be arranged to communicate with a storage medium 930 to execute a series of instruction operations in the storage medium 930 on the training device 900.

The training device 900 may also include one or more power sources 926, one or more wired or wireless network interfaces 950, one or more input/output interfaces 958; or one or more operating systems 941, such as Windows ServerTM, mac OS XTM, unixTM, linuxTM, freeBSDTM, etc.

Specifically, the training device may perform the steps 401 to 404 in the above embodiments.

Embodiments of the present application also provide a computer program product that, when run on a computer, causes the computer to perform the steps performed by the aforementioned performing device, or causes the computer to perform the steps performed by the aforementioned training device.

There is also provided in an embodiment of the present application a computer-readable storage medium having stored therein a program for performing signal processing, which when run on a computer, causes the computer to perform the steps performed by the aforementioned performing device or causes the computer to perform the steps performed by the aforementioned training device.

The execution device, training device or terminal device provided in the embodiment of the present application may specifically be a chip, where the chip includes: a processing unit, which may be, for example, a processor, and a communication unit, which may be, for example, an input/output interface, pins or circuitry, etc. The processing unit may execute the computer-executable instructions stored in the storage unit to cause the chip in the execution device to perform the data processing method described in the above embodiment, or to cause the chip in the training device to perform the data processing method described in the above embodiment. Optionally, the storage unit is a storage unit in the chip, such as a register, a cache, etc., and the storage unit may also be a storage unit in the wireless access device side located outside the chip, such as a read-only memory (ROM) or other type of static storage device that may store static information and instructions, a random access memory (random access memory, RAM), etc.

Specifically, referring to fig. 10, fig. 10 is a schematic structural diagram of a chip provided in an embodiment of the present application, where the chip may be represented as a neural network processor NPU1000, and the NPU1000 is mounted as a coprocessor on a main CPU (Host CPU), and the Host CPU distributes tasks. The core part of the NPU is an arithmetic circuit 1003, and the controller 1004 controls the arithmetic circuit 1003 to extract matrix data in the memory and perform multiplication.

The NPU1000 may implement the information data processing method provided in the embodiment depicted in fig. 4 through inter-cooperation between the various devices within.

More specifically, in some implementations, the arithmetic circuit 1003 in the NPU1000 includes a plurality of processing units (PEs) internally. In some implementations, the operational circuit 1003 is a two-dimensional systolic array. The arithmetic circuit 1003 may also be a one-dimensional systolic array or other electronic circuitry capable of performing mathematical operations such as multiplication and addition. In some implementations, the arithmetic circuit 1003 is a general purpose matrix processor.

For example, assume that there is an input matrix a, a weight matrix B, and an output matrix C. The arithmetic circuit takes the data corresponding to matrix B from the weight memory 1002 and buffers it on each PE in the arithmetic circuit. The arithmetic circuit takes matrix a data from the input memory 1001 and performs matrix operation with matrix B, and the obtained partial result or final result of the matrix is stored in an accumulator (accumulator) 1008.

The unified memory 1006 is used for storing input data and output data. The weight data is directly transferred to the weight memory 1002 through the memory cell access controller (Direct Memory Access Controller, DMAC) 1005. The input data is also carried into the unified memory 1006 through the DMAC.

BIU is Bus Interface Unit, bus interface unit 1010, for the AXI bus to interact with DMAC and finger memory (Instruction Fetch Buffer, IFB) 1009.

The bus interface unit 1010 (Bus Interface Unit, abbreviated as BIU) is configured to obtain an instruction from the external memory by the instruction fetch memory 1009, and further configured to obtain the raw data of the input matrix a or the weight matrix B from the external memory by the memory unit access controller 1005.

The DMAC is mainly used to transfer input data in the external memory DDR to the unified memory 1006 or to transfer weight data to the weight memory 1002 or to transfer input data to the input memory 1001.

The vector calculation unit 1007 includes a plurality of operation processing units that perform further processing such as vector multiplication, vector addition, exponential operation, logarithmic operation, size comparison, and the like on the output of the operation circuit 1003, if necessary. The method is mainly used for non-convolution/full-connection layer network calculation in the neural network, such as Batch Normalization (batch normalization), pixel-level summation, up-sampling of a characteristic plane and the like.

In some implementations, the vector calculation unit 1007 can store the vector of processed outputs to the unified memory 1006. For example, the vector calculation unit 1007 may be configured to perform a linear function; alternatively, a nonlinear function is applied to the output of the arithmetic circuit 1003, such as linear interpolation of the feature planes extracted by the convolutional layer, and then such as a vector of accumulated values, to generate the activation value. In some implementations, the vector calculation unit 1007 generates a normalized value, a pixel-level summed value, or both. In some implementations, the vector of processed outputs can be used as an activation input to the arithmetic circuit 1003, for example for use in subsequent layers in a neural network.

An instruction fetch memory (instruction fetch buffer) 1009 connected to the controller 1004, for storing instructions used by the controller 1004;

the unified memory 1006, the input memory 1001, the weight memory 1002, and the finger memory 1009 are all On-Chip memories. The external memory is proprietary to the NPU hardware architecture.

The processor mentioned in any of the above may be a general-purpose central processing unit, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the above-mentioned programs.

It should be further noted that the above-described apparatus embodiments are merely illustrative, and that the units described as separate units may or may not be physically separate, and that units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the embodiment of the device provided by the application, the connection relation between the modules represents that the modules have communication connection therebetween, and can be specifically implemented as one or more communication buses or signal lines.

From the above description of the embodiments, it will be apparent to those skilled in the art that the present application may be implemented by means of software plus necessary general purpose hardware, or of course may be implemented by dedicated hardware including application specific integrated circuits, dedicated CPUs, dedicated memories, dedicated components and the like. Generally, functions performed by computer programs can be easily implemented by corresponding hardware, and specific hardware structures for implementing the same functions can be varied, such as analog circuits, digital circuits, or dedicated circuits. However, a software program implementation is a preferred embodiment in many cases for the present application. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a readable storage medium, such as a floppy disk, a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk or an optical disk of a computer, etc., including several instructions for causing a computer device (which may be a personal computer, a training device, or a network device, etc.) to perform the method described in the embodiments of the present application.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product.

The computer program product includes one or more computer instructions. When loaded and executed on a computer, produces a flow or function in accordance with embodiments of the present application, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via a wired (e.g., coaxial cable, optical fiber, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be stored by a computer or a data storage device such as a training device, a data center, or the like that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., a floppy Disk, a hard Disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.

Claims

1. A method of data processing, the method comprising:

acquiring a first feature representation and a second feature representation; the first characteristic is obtained based on attribute information of the user and the article and first operation information of the user on the article, the first operation information indicates whether the first operation behavior exists on the article by the user, and the second characteristic is obtained based on attribute information of the user and the article and second operation information of the user on the article; the second operation information indicates whether a second operation action exists on the article by the user, and the first operation action is an associated action on the article before the user performs the second operation action;

and updating the first characteristic representation and the second characteristic representation according to the first prediction result.

2. The method of claim 1, wherein the obtaining the feature representation of the first feature representation that is related to the second feature representation is a third feature representation, comprising:

3. A method according to claim 1 or 2, characterized in that,

4. A method according to any one of claims 1 to 3, wherein the method further comprises: predicting a second prediction result of the operation behavior of the user on the article according to the second characteristic representation;

said updating said first and second feature representations according to said first prediction result, comprising:

5. The method according to any one of claims 1 to 4, further comprising: predicting a third prediction result of the operation behavior of the user on the article according to the first characteristic representation;

6. The method according to any one of claims 1 to 5, further comprising: acquiring a feature representation which is irrelevant to the second feature representation in the first feature representation as a fourth feature representation; acquiring a feature representation related to the first feature representation in the fourth feature representation as a fifth feature representation;

7. The method of claim 6, wherein the obtaining a feature representation of the first feature representation that is not related to the second feature representation is a fourth feature representation, comprising:

8. The method according to claim 6 or 7, characterized in that the method further comprises:

9. The method according to any one of claims 1 to 8, further comprising:

10. The method according to claim 9, wherein the method further comprises: predicting a sixth prediction result of the user's presence of an operational behavior on the item according to the sixth feature representation;

11. A data processing apparatus, the apparatus comprising:

12. The apparatus according to claim 11, wherein the processing module is specifically configured to:

13. The device according to claim 11 or 12, wherein,

14. The apparatus of any one of claims 11 to 13, wherein the processing module is further configured to:

the updating module is specifically configured to:

15. The apparatus of any one of claims 11 to 14, wherein the processing module is further configured to:

the updating module is specifically configured to:

16. The apparatus of any one of claims 11 to 15, wherein the processing module is further configured to:

the updating module is specifically configured to:

17. The apparatus according to claim 16, wherein the processing module is specifically configured to:

18. The apparatus of claim 16 or 17, wherein the processing module is further configured to:

The updating module is specifically configured to:

19. The apparatus of any one of claims 11 to 18, wherein the processing module is further configured to:

20. The apparatus of claim 19, wherein the processing module is further configured to:

the updating module is specifically configured to:

21. A computing device, the computing device comprising a memory and a processor; the memory stores code, the processor being configured to retrieve the code and perform the method of any of claims 1 to 10.

22. A computer storage medium storing one or more instructions which, when executed by one or more computers, cause the one or more computers to implement the method of any one of claims 1 to 10.

23. A computer program product comprising code for implementing the method of any of claims 1 to 10 when said code is executed.