CN116992116A

CN116992116A - Content recommendation method, device, electronic equipment, storage medium and program product

Info

Publication number: CN116992116A
Application number: CN202210433319.8A
Authority: CN
Inventors: 胡乐
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2022-04-24
Filing date: 2022-04-24
Publication date: 2023-11-03

Abstract

The embodiment of the application discloses a content recommendation method, a device, electronic equipment, a storage medium and a program product. The method comprises the following steps: acquiring operation time, operation data and an operation object corresponding to the operation data of a target user; acquiring an object attribute corresponding to the operation object; according to the object attribute, determining a candidate operation object corresponding to the target user; calculating a target weight value of each candidate operation object according to the operation time and the object attribute; based on the target weight value, the method can integrate the influence of the interests of the user on the recommendation from the candidate operation objects to ensure that the target operation objects recommended to the target user are related to the interests of the user, thereby improving the accuracy of the recommendation.

Description

Content recommendation method, device, electronic equipment, storage medium and program product

Technical Field

The present application relates to the field of computer technologies, and in particular, to a content recommendation method, apparatus, electronic device, storage medium, and program product.

Background

The Internet brings a large amount of information to people, meets the information demand of people in the information age, and can not obtain information really useful for the people when the people face a large amount of information. The recommendation system can predict the information preference degree of the target user, provide the most interesting content for the user from massive information and push the content to the user so that the user can quickly acquire information useful for the user.

However, when the current recommendation system recommends content, the value of the content is generally measured according to the flow value of the content, and the influence of the actual operation of the user on the recommended content is ignored, so that the accuracy of content recommendation is low.

Disclosure of Invention

The embodiment of the application provides a content recommendation method, a content recommendation device, electronic equipment, a storage medium and a program product, which can improve the accuracy of content recommendation.

The embodiment of the application provides a content recommendation method, which comprises the following steps:

acquiring operation time, operation data and an operation object corresponding to the operation data of a target user;

acquiring an object attribute corresponding to the operation object;

according to the object attribute, determining a candidate operation object corresponding to the target user;

Calculating a target weight value of each candidate operation object according to the operation time and the object attribute;

and determining a target operation object recommended to the target user from the candidate operation objects based on the target weight value.

The embodiment of the application also provides a content recommendation device, which comprises:

the first acquisition module is used for acquiring the operation time and the operation data of the target user and an operation object corresponding to the operation data;

the second acquisition module is used for acquiring the object attribute corresponding to the operation object;

the candidate determining module is used for determining candidate operation objects corresponding to the target users according to the object attributes;

the weight calculation module is used for calculating a target weight value of each candidate operation object according to the operation time and the article attribute;

and the recommending module is used for determining a target operation object recommended to the target user from the candidate operation objects based on the target weight value.

In some embodiments, the recommendation module further comprises:

the profit score obtaining unit is used for obtaining the profit score corresponding to the candidate operation object;

the calculation unit is used for carrying out weighted calculation on the benefit score based on the target weight value to obtain a target score;

And a determining unit configured to determine a target operation object recommended to the target user from among the candidate operation objects based on the target score.

In some embodiments, the item attribute comprises a plurality of sub-item attributes, the target weight value comprises a first weight value, a second weight value, and a third weight value, and the weight calculation module further comprises:

a first weight determining unit, configured to determine a first weight value according to the attribute of the child object used to obtain the candidate operation object;

a second weight determining unit, configured to determine a second weight value of the candidate operation object according to the operation data corresponding to the child object attribute;

and a third weight calculation unit for calculating the third weight value according to the difference value between the operation time and the current time.

In some embodiments, the item attribute comprises a plurality of sub-item attributes, and the candidate determination module further comprises:

a sub-candidate determining unit, configured to determine a sub-candidate operation object corresponding to each sub-item attribute according to each sub-item attribute;

and the candidate determining unit is used for determining all the sub candidate operation objects as the candidate operation objects.

In some embodiments, the sub-candidate determination unit is further to:

Acquiring embedded representations corresponding to the attributes of each sub-item;

calculating the similarity between any two sub-item attributes according to the embedded representation corresponding to each sub-item attribute;

determining a target sub-item attribute corresponding to each sub-item attribute according to the similarity;

and determining the operation object corresponding to the target sub-item attribute as a sub-candidate operation object corresponding to the sub-item attribute.

In some embodiments, the sub-candidate determination unit is further to:

acquiring a corresponding relation between a preset sub-item attribute and the operation object, wherein the preset sub-item attribute comprises the sub-item attribute;

and determining the operation object corresponding to each sub-item attribute according to the corresponding relation to obtain a sub-candidate operation object corresponding to each sub-item attribute.

The embodiment of the application also provides electronic equipment, which comprises a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute steps in any content recommendation method provided by the embodiment of the application.

The embodiment of the application also provides a computer readable storage medium, which stores a plurality of instructions, the instructions are suitable for being loaded by a processor to execute the steps in any content recommendation method provided by the embodiment of the application.

The embodiments of the present application also provide a computer program product comprising computer programs/instructions which, when executed by a processor, implement the steps in any of the video processing methods provided by the embodiments of the present application.

According to the method and the device for recommending the target operation object, the operation time, the operation data and the operation object of the target user can be obtained, the object attribute corresponding to the operation object is obtained, the candidate operation object which the target user is likely to be interested in is determined by using the object attribute, different target weight values are given to the candidate operation object based on the object attribute and the operation time, the influence of the interests of the user is fully considered, the fact that the target operation object recommended to the target user is related to the interests of the user is guaranteed, and therefore recommending accuracy is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1a is a schematic view of a scenario of a content recommendation method according to an embodiment of the present application;

FIG. 1b is a flowchart illustrating a content recommendation method according to an embodiment of the present application;

FIG. 1c is a schematic diagram of a relationship between data to be processed and properties of an article according to an embodiment of the present application;

FIG. 2 is a schematic diagram of a content recommendation method applied in an advertisement recommendation scene according to an embodiment of the present application;

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

fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to fall within the scope of the application.

The embodiment of the application provides a content recommendation method, a content recommendation device, electronic equipment and a storage medium.

The content recommendation device may be integrated in an electronic device, and the electronic device may be a terminal, a server, or other devices. The terminal can be a mobile phone, a tablet personal computer, an intelligent Bluetooth device, an intelligent voice interaction device, an intelligent household appliance, a vehicle-mounted terminal, an aircraft, a notebook computer, a personal computer (Personal Computer, PC) or the like; the server may be a single server or a server cluster composed of a plurality of servers.

In some embodiments, the content recommendation device may also be integrated in a plurality of electronic devices, for example, the content recommendation device may be integrated in a plurality of servers, and the content recommendation method of the present application is implemented by the plurality of servers.

In some embodiments, the server may also be implemented in the form of a terminal.

For example, referring to fig. 1a, an application scenario schematic diagram of a content recommendation method provided by an embodiment of the present application is shown.

As shown in fig. 1a, the user terminal 101 and the server 102 are located in a wireless network or a wired network, and the user terminal 101 and the server 102 perform data interaction. The server 102 may be a separate server, may be a server cluster, may be a local server, or may be a cloud server.

The server 102 is capable of acquiring data to be processed through the user terminal 101. The data to be processed refers to data related to a target user, and may include operation time, operation data, and an operation object corresponding to the operation data.

The operation time may be time information corresponding to when the target user performs a certain operation on the operation object, for example, when the target user clicks a certain operation object at the point 08 of 2 nd month 3 of 2022, the operation time is then the point 08 of 3 nd 2 nd year 2022; the manner in which the server 102 obtains the operational data may be by the targeted user actively reporting or by SDK components within various applications within the user terminal 101.

The operation data may refer to data generated when the target user performs an operation at the user terminal 101, and the operation data may include a specific operation type, an operation duration, and the like, for example, the target user browses a certain advertisement for 2 minutes, the target user browses a certain advertisement for 3 times, and the like; the manner in which the server 102 obtains the operational data may be by the targeted user actively reporting or by SDK components within various applications within the user terminal 101.

The operation object corresponding to the operation data may refer to information of an object acted upon by the operation of the target user, for example, information of a video viewed by the target user, information of a clicked advertisement, and the like; the manner in which the server 102 obtains the operational data may be by the targeted user actively reporting or by SDK components within various applications within the user terminal 101.

The server 102 may obtain the operation time, the operation data, the operation object corresponding to the operation data, and the object attribute corresponding to the operation object of the target user; according to the object attribute, determining a candidate operation object corresponding to the target user; then, according to the operation time and the object attribute, a target weight value of each candidate operation object is calculated, and then the target weight value is used to determine a target operation object from the candidate operation objects, and the target operation object is pushed to the user terminal 101.

In some embodiments, the server 102 may be used as a node in a blockchain system, or may implement the above steps, for example, after the operation time, the operation data, and the operation object corresponding to the operation data of the target user are acquired, the operation data may be stored in a common uplink through a blockchain, so as to ensure the authenticity of the acquired data, and prevent the data from being tampered, so that unrealistic data or tampered data may be avoided for content recommendation, and thus the accuracy of content recommendation may be improved.

The blockchain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. The Blockchain (Blockchain), which is essentially a decentralised database, is a string of data blocks that are generated by cryptographic means in association, each data block containing a batch of information of network transactions for verifying the validity of the information (anti-counterfeiting) and generating the next block. The blockchain may include a blockchain underlying platform, a platform product services layer, and an application services layer.

The blockchain underlying platform may include processing modules for user management, basic services, smart contracts, and operational monitoring. The user management module is responsible for identity information management of all blockchain participants, including maintenance of public and private key generation (account management), key management, maintenance of corresponding relation between the real identity of the user and the blockchain address (authority management) and the like, and under the condition of authorization, supervision and audit of transaction conditions of certain real identities, and provision of rule configuration (wind control audit) of risk control; the basic service module is deployed on all block chain node devices, is used for verifying the validity of a service request, recording the service request on a storage after the effective request is identified, for a new service request, the basic service firstly analyzes interface adaptation and authenticates the interface adaptation, encrypts service information (identification management) through an identification algorithm, and transmits the encrypted service information to a shared account book (network communication) in a complete and consistent manner, and records and stores the service information; the intelligent contract module is responsible for registering and issuing contracts, triggering contracts and executing contracts, a developer can define contract logic through a certain programming language, issue the contract logic to a blockchain (contract registering), invoke keys or other event triggering execution according to the logic of contract clauses to complete the contract logic, and simultaneously provide a function of registering contract upgrading; the operation monitoring module is mainly responsible for deployment in the product release process, modification of configuration, contract setting, cloud adaptation and visual output of real-time states in product operation, for example: alarming, monitoring network conditions, monitoring node equipment health status, etc.

The platform product service layer provides basic capabilities and implementation frameworks of typical applications, and developers can complete the blockchain implementation of business logic based on the basic capabilities and the characteristics of the superposition business. The application service layer provides the application service based on the block chain scheme to the business participants for use.

It should be noted that, in the embodiment of the present application, data related to the target user, such as operation time, operation data, an operation object corresponding to the operation data, and other data related to the target user, such as a duration of use of a certain application program by the target user, etc., when the embodiment of the present application applies these data to a specific product or technology, user permission or consent needs to be obtained, and collection, use and processing of related data need to comply with related laws and regulations and standards of related countries and regions, respectively, as described in detail below.

Artificial intelligence (Artificial Intelligence, AI) is a technology that utilizes a digital computer to simulate the human perception environment, acquire knowledge, and use the knowledge, which can enable machines to function similar to human perception, reasoning, and decision. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning, automatic driving, intelligent traffic and other directions.

Machine Learning (ML) is a multi-domain interdisciplinary, involving multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. It is specially studied how a computer simulates or implements the learning characteristics of a human being to acquire new knowledge or skills, reorganizing the existing knowledge structure to continuously improve its own performance. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like.

With research and progress of artificial intelligence technology, research and application of artificial intelligence technology are being developed in various fields, such as common smart home, smart wearable devices, virtual assistants, smart speakers, smart marketing, unmanned, autopilot, unmanned, robotic, smart medical, smart customer service, car networking, autopilot, smart transportation, etc., and it is believed that with the development of technology, artificial intelligence technology will be applied in more fields and will be of increasing importance.

In this embodiment, a content recommendation method related to artificial intelligence is provided, as shown in fig. 1b, the specific flow of the content recommendation method may be as follows:

s110, acquiring operation time, operation data and an operation object corresponding to the operation data of a target user.

The target user is a user needing content recommendation, and when the content recommendation is performed on the target user, the data to be processed of the target user can be acquired first, wherein the data to be processed can comprise operation time, operation data and an operation object corresponding to the operation data.

The operation time may be time information corresponding to when the target user performs a certain operation on the operation object, for example, when the target user clicks a certain operation object at point 08 of 2022, 2 and 3, then the operation time is at point 08 of 2022, 2 and 3.

The operation data may refer to data generated when the target user performs an operation at the user terminal 101, and the operation data may include a specific operation and parameters related to the operation, etc., for example, the target user browses a certain advertisement for 2 minutes, the target user browses a certain advertisement for 3 times, the target user clicks a certain book for 3 times, etc.

The operation object corresponding to the operation data may refer to information of an object on which the operation of the target user is performed, for example, information of a video viewed by the target user, information of a clicked advertisement, and the like. From the above, it is known that the operation time, the operation data and the operation object have an association relationship, that is, the corresponding operation time, operation data and operation object can be generated by one operation of the target user.

It should be noted that, the operation time, the operation data, and the operation object corresponding to the operation data of the target user are all obtained under the authorization of the target user, or are actively transmitted by the target user.

In some embodiments, the data to be processed may be acquired at regular intervals, for example, the data to be processed of the target user is acquired once every preset time interval, if the authorization of the target user is obtained.

In some embodiments, the data to be processed may be acquired when the triggering operation of the user is detected in the case of being authorized by the target user. The triggering operation may be an interface refreshing operation, an operation of entering an application program, or the like, and may be specifically set according to actual needs, which is not specifically limited herein.

S120, acquiring the object attribute corresponding to the operation object.

After the operation object corresponding to the operation data is acquired, the item attribute corresponding to the operation object may be generated. Wherein, the operation object may generally specifically point to a certain item, for example, when the operation object is an advertisement, the advertisement generally corresponds to the certain item, and the item attribute may include a plurality of sub-item attributes, and the sub-item attribute may include an attribute of an item corresponding to the operation object and an identifier of the operation object.

In some embodiments, the attribute of the object may be a specific category including the object to which the operation object corresponds, and the specific category to which the object belongs may include a first category, a second category, a third category, and so on, and then the attribute of the object may refer to the attribute of the object under a specific category. For example, the article pointed by a certain advertisement is a mobile phone of a Z brand, then the article pointed by the advertisement is a mobile phone, the attribute of the mobile phone in a first class can be a digital product, the attribute of the mobile phone in a second class can be a mobile phone, and the attribute of the mobile phone in a third class can be a Z brand.

The identification of the operation object may refer to the name of the operation object, or the identification of the operation object in the recommendation database, where the identification of the operation object may refer to the operation object uniquely in the recommendation database. The recommendation database may be used to store all recommended operation objects, which may be advertisements, videos, music, books, commodities, web pages, etc., and may be set according to actual needs, which are not limited herein.

As can be seen from the foregoing description, the object attributes corresponding to the operation object may include a plurality of sub-object attributes, where the sub-object attributes are an attribute under the first class, an attribute under the second class, an attribute under the third class, and an identifier of the operation object.

When the object attribute corresponding to the operation object is acquired, a plurality of sub-object attributes may be acquired. For example, the identifier of the operation object may be obtained according to the obtained operation object, then the item corresponding to the operation object is determined, and the category of the item is determined, so that the item attribute corresponding to the operation object may be obtained.

In some embodiments, the object attribute corresponding to each operation object in the recommendation database may be predetermined, and stored in a preset position, so that the preset position may be directly accessed, and the object attribute corresponding to the operation object may be obtained.

In some embodiments, the object attribute may be generated in real time according to the actual operation of the target user, and the object attribute and the data to be processed are stored together, so that the data to be processed and the corresponding object attribute may be obtained simultaneously. For example, referring to fig. 1c, a schematic diagram of a relationship between data to be processed and an item attribute is shown, wherein the item attribute is determined according to an operation object, and the operation data, the operation object and the operation time have an association relationship, so that the item attribute also has an association relationship with the operation data, the operation object and the operation time.

S130, determining candidate operation objects corresponding to the target users according to the object attributes.

After the object attribute is obtained, a candidate operation object corresponding to the target user can be determined according to the object attribute, wherein the candidate operation object can be an operation object which may be of interest to the target user.

The object attribute is related to the operation of the target user, and the interest of the target user can be reflected, so that the candidate operation object can be determined by using the object attribute. In some embodiments, the item attribute may include a plurality of sub-item attributes, and when determining the candidate operation object according to the item attribute, a sub-candidate operation object corresponding to each sub-item attribute may be determined based on the sub-item attribute, and then sub-candidate operation objects corresponding to all sub-item attributes are determined as candidate operation objects.

In some embodiments, in order to reduce the data processing amount during recommendation, the preset sub-item attribute and the operation object may be bound in advance, so as to obtain the corresponding relationship between the preset sub-item attribute and the operation object. When determining the sub-candidate operation object corresponding to the sub-object attribute, the corresponding relation between the preset sub-object attribute and the operation object can be directly obtained, and the operation object corresponding to each sub-object attribute is determined according to the corresponding relation between the preset sub-object attribute and the operation object, so as to obtain the sub-candidate operation object corresponding to each sub-object attribute.

The preset sub-item attribute refers to all possible item attributes, for example, may be item attributes corresponding to all operation objects in the recommendation data, so that it is known that the preset sub-item attribute may include a sub-item attribute.

As an implementation manner, before the recommendation, the preset sub-object attributes are bound with the operation object, so as to obtain the corresponding relationship between the preset sub-object attributes and the operation object. When the preset sub-object attributes and the operation objects are bound, an embedding model can be obtained through training in advance, embedding representations corresponding to the preset sub-object attributes are obtained through the embedding model, then the embedding representations corresponding to any two preset sub-object attributes are multiplied to obtain the similarity between any two preset sub-object attributes, then the target preset sub-object attribute corresponding to each preset sub-object attribute is determined according to the similarity, and then the operation objects corresponding to the target preset sub-object attributes and the preset sub-object attributes are bound to obtain the corresponding relation between the preset sub-object attributes and the operation objects.

Therefore, after the corresponding relation between the preset sub-object attribute and the operation object is obtained, the operation object corresponding to the sub-object attribute can be found, and the sub-candidate operation object is obtained.

In some embodiments, in order to improve accuracy of recommendation, real-time calculation may be performed during recommendation, and when determining sub-candidate operation objects corresponding to sub-item attributes, an embedded representation corresponding to each sub-item attribute may be obtained; according to the embedded representation corresponding to each sub-item attribute, calculating the similarity between any two sub-item attributes, determining a target sub-item attribute corresponding to each sub-item attribute, and determining an operation object corresponding to the target sub-item attribute as a sub-candidate operation object corresponding to the sub-item attribute.

The embedded representation corresponding to the child object attribute can be obtained through a trained embedded model, and the embedded model refers to a model capable of generating the embedded representation, namely, a word is represented as a vector, for example, when a word is input into the embedded model, the embedded model can output the embedded representation of the word, and the embedded model is an important technology in natural language processing. The embedded model may be a neural network based embedded model; for example, the embedding model may include: word2vec, etc.

Word2vec can map words into a low-dimensional vector space, and the similarity between words can be obtained by calculating the distance between two words. The Word2vec model type may be various, and may include, for example, a continuous Word bag (Continuous Bag of Words, CBOW) model, a Skip-gram model, and so on.

Before using the embedded model, the embedded model needs to be trained using a sample sequence, which may be obtained using all the objects operated by all the users during the history period. The historical period refers to a preset period of time before the current time, for example, within 7 days before the current time, and within 1 day before the current time, where the preset period of time may be set according to actual needs, and is not specifically limited herein. All users refer to users who can make content recommendation, for example, in a certain preset application program, and then all users may refer to users who have account numbers in the preset application program.

It can be understood that all the operation objects operated by the user in the history period have their corresponding object attributes, and the object attributes include sub-object attributes, where the sub-object attributes may be the identifiers of the operation objects and the attributes under each category, and the constructed sample sequence may be a sequence of constructing the identifiers of the operation objects for each user, a sequence of the attributes under each category, and a mixed sequence of the identifiers of the operation objects and the attributes under each category, that is, the corresponding attributes are sequentially arranged according to the time sequence, so as to obtain the corresponding sequence. For example, when a user clicks on an operation object identified as A1, A2, A3 in turn, the sequence of the identifications of the operation objects constructed may be [ A1, A2, A3]. Model training is carried out through the sample sequences, so that an embedded model can be obtained, and by using the embedded model, embedded representations corresponding to the attributes of the child objects can be obtained.

After the embedded representation corresponding to each sub-item attribute is obtained, similarity calculation can be performed according to the embedded representation corresponding to each sub-item attribute, so as to further determine sub-candidate operation objects corresponding to the sub-item attributes according to the similarity. When similarity calculation is performed based on the embedded representation of the object attribute, the situation that when a new operation object is added into the recommendation database, the new operation object cannot be associated with the interest of the user can be avoided, and each operation object can be ensured to be associated with the user.

In some embodiments, the sub-item attributes may include identifiers of the operation objects and attributes under different classes, if the sub-item attributes are identifiers of the operation objects, the identifiers of all the operation objects in the recommendation database may be obtained, the embedded representations of the sub-item attributes are multiplied by the embedded representations of the identifiers of all the operation objects to obtain a similarity corresponding to each operation object identifier, the identifiers of the topK operation objects are filtered according to the similarity to obtain a target sub-item attribute corresponding to the sub-item attribute, and the operation object corresponding to the target sub-item attribute is determined to be a sub-candidate operation object corresponding to the sub-item attribute.

Similarly, if the attribute of the sub-item is an attribute under the first class, the attribute under the first class corresponding to all the operation objects in the recommendation database may be obtained, and the similarity is calculated according to the above manner, so as to obtain the sub-candidate operation object corresponding to each sub-item attribute.

It should be noted that, when determining the corresponding relationship between the attribute of the preset sub-object and the operation object, the method is similar to the method, and in order to avoid repetition, the description is omitted here.

And S140, calculating the target weight value of each candidate operation object according to the operation time and the object attribute.

The number of candidate operation objects is usually a plurality of, and in order to further determine the operation object most interested by the user, a target weight value corresponding to each candidate operation object may be calculated according to the operation time and the object attribute.

In some implementations, the target weight value may include a first weight value, a second weight value, and a third weight value. The first weight value may be a weight value obtained according to an attribute of the article, the second weight value may be a weight value obtained according to operation data, and the third weight value may be a weight value obtained according to operation time.

In some embodiments, the first weight value may be determined based on a child property used to derive the candidate operand. As described above, the candidate operation object is obtained based on the child item attribute, and the child candidate operation objects corresponding to different child item attributes may be the same, so that one candidate operation object may correspond to a plurality of different child item attributes. Thus, the sub-item attributes corresponding to the same candidate operation object may be aggregated, for example, all sub-item attributes used by the candidate operation object are determined to be acquired, recommended granularity corresponding to the sub-item attributes is acquired, and the sub-item attribute with the highest recommended granularity is reserved.

As described above, the sub-item attribute may refer to an identifier of the operation object, a first category of the item, a second category of the item, and a third category of the item, and the recommended granularity of these sub-item attributes may be preset, for example, the recommended granularity corresponding to the identifier of the operation object may be set to be the highest, the recommended granularity corresponding to the third category of the item is smaller than the recommended granularity corresponding to the identifier of the operation object, the recommended granularity corresponding to the second category of the item is smaller than the recommended granularity corresponding to the third category, and the recommended granularity corresponding to the first category of the item is smaller than the recommended granularity corresponding to the second category. If the sub-item attribute corresponding to a certain candidate operation object comprises the identifier under the second class and the identifier of the operation object, and the recommended granularity corresponding to the identifier of the operation object is larger than the recommended granularity corresponding to the second class, the sub-item attribute corresponding to the candidate operation object can be combined into the identifier of the operation object.

It can be understood that after merging the sub-item attributes of each candidate operation object, one candidate operation object corresponds to one type of sub-item attribute, so that a weight value corresponding to the sub-item attribute can be determined, and the first weight value is obtained. That is, after the sub-item attributes are combined, each candidate operation object only has any one of the identification of the operation object, the attribute under the primary class, the attribute under the secondary class and the attribute under the tertiary class.

In some embodiments, when determining the first weight value, the corresponding relationship between the type of the attribute of the child item and the weight value may be obtained, a target type corresponding to the attribute of the child item is determined, and based on the corresponding relationship, the weight value corresponding to the target type is determined, so as to obtain the first weight value. For example, referring to table 1, correspondence of types of child item attributes and weight values is shown.

TABLE 1

Type(s)	Identification of an operation object	Three-level category	Second class of	First class of
					Weight value	x1	x2	x3	x4

In table 1, each type of child object attribute corresponds to a weight value thereof, where the weight value x1> x2> x3> x4, and the first weight value corresponding to the candidate operation object can be obtained through the correspondence shown in table 1.

In some embodiments, the second weight value of the candidate operation object may be determined according to the operation data corresponding to the child object attribute. The operation data corresponding to the child item attribute is operation data corresponding to an operation when the child item attribute is generated, and it is to be noted that the child item attribute is generated according to an operation object, and the operation object is an object acted on by the operation data, so that the child item attribute and the operation data have a corresponding relationship, and the operation data corresponding to the child item attribute can be obtained.

In some embodiments, when determining the second weight value, the target operation type may be determined according to the operation data, and the weight value corresponding to the target operation type may be determined according to the correspondence between the operation type and the weight value, so as to obtain the second weight value.

When determining the target operation type, a specific operation may be obtained from the operation data, an interest feature value corresponding to the specific operation is determined, and a corresponding operation type is determined. The interest degree value is used for indicating the interest degree of a target user on a certain operation object or article, and interest characteristic values reflected by different operations are different. The correspondence between the specific operation and the interest characteristic value may be stored in advance, so that the interest characteristic value corresponding to each operation may be obtained. For example, the interest feature values reflected by the clicking operation and the browsing operation are the same, the interest degree value corresponding to the clicking operation is 1, the interest degree value corresponding to the viewing operation is 1, the interest degree value corresponding to the downloading operation is 2, and the like, so that the interest feature value corresponding to each specific operation can be obtained. And mapping the interest characteristic value to the corresponding operation type according to the corresponding relation between the interest characteristic value and the operation type, thereby obtaining the target operation type. For example, the click operation and the browse operation may be a first type of operation, the download operation may be a second type of operation, the install operation may be a third type of operation, the activate operation and the register operation may be a fourth type of operation, the pay operation may be a fifth type of operation, and the pay operation may be specifically set according to actual needs, which is not specifically limited herein.

For example, table 2 may be referred to, showing correspondence of operation types and weight values.

TABLE 2

Operation type	Of the first type	Of the second type	Third type	Type IV	Of the fifth type
						Weight value	y1	y2	y3	y4	y5

In table 2, each operation type corresponds to its weight value, where the weight value y1< y2< y3< y4< y5, and the second weight value corresponding to the candidate operation object can be obtained through the correspondence shown in table 2.

In some embodiments, the third weight value may be determined according to the operation time, for example, a time difference between the operation time and the current time may be calculated, and the third weight value may be calculated according to the time difference.

As an embodiment, the third weight value may be calculated according to the following formula:

wherein, weihht _t The third weight value is represented, t represents the difference between the operation time and the current time, and according to the calculation formula, it can be seen that the larger the difference between the operation time and the current time is, the lower the third weight value is, which represents that the interest of the user has less influence on the current time.

After the first weight value, the second weight value, and the third weight value are acquired, it may be considered that the target weight value is acquired.

In some embodiments, the candidate operation object may have more than one sub-item attribute, and the foregoing merely ensures that the candidate operation object corresponds to one type of sub-item attribute, and then the candidate operation object may have multiple sub-item attributes of one type. For example, the candidate operation object corresponds to the child item attribute under the two three-level classes, namely child item attribute 1, and child item attribute 2. Since the target weight value and the child item attribute have an association relationship, one candidate operation object may correspond to a plurality of target weight values.

As in the foregoing example, the sub-item attributes corresponding to the candidate operation object are sub-item attribute 1 and sub-item attribute 2, which are sub-item attributes under the three-level category, it may be understood that the first weight values corresponding to the sub-item attribute 1 and the sub-item attribute 2 are the same, and taking the correspondence relationship shown in the foregoing table 1 as an example, it may be determined that the first weight values are x2, but the corresponding second weight values and third weight values may be different.

The second weight value is determined according to the operation data corresponding to the child item attribute, and then the operation data corresponding to the child item attribute 1 and the child item attribute 2 may be different. For example, sub-item attribute 1 is generated when the target user downloads a certain application program, and sub-item attribute 2 is generated when the target user clicks a certain application program, then two second weight values exist for the candidate operation object, and it is assumed that the second weight values are y1 and y2, respectively, where y1 corresponds to sub-item attribute 1, and y2 corresponds to sub-item attribute 2.

If the operation data corresponding to the sub-item attribute 1 and the sub-item attribute 2 are the same, the corresponding second weight value is the same.

The third weight value is determined according to the operation time, and the operation time, the operation data and the sub-item attributes have an association relationship, and the operation times corresponding to the different sub-item attributes are different, so that the candidate operation object has two third weight values, such as weight _t 1 and weight _t 2, wherein weight _t 1 child item attribute 1 corresponds to weight _t 2 corresponds to child item attribute 2.

From the foregoing, it can be seen that the first weight value, the second weight value, and the third weight value included in the target weight value have an association relationship. Then, when the candidate operation object corresponds to the child object attribute 1 and the child object attribute 2, two target weight values are corresponding, as described above, wherein one target weight value may be x2, y1, and weight _t 1, another target weight value is x2, y2 and weight _t 2。

And S150, determining a target operation object recommended to the target user from the candidate operation objects based on the target weight value.

The target operation object is an operation object recommended to the target user, and after the target weight value is obtained, the target operation object can be determined from the candidate operation objects based on the target weight value.

In some embodiments, a target score corresponding to each candidate operation object may be calculated according to the target weight value, and the target operation object is determined from the candidate operation objects based on the target score. In order to avoid excessive waste of the flow value of the operation object, the flow value of the candidate operation object may be quantized to obtain a benefit score corresponding to the candidate operation object, and the benefit score may be weighted based on the target weight value to obtain the target score.

The benefit score may refer to a benefit (effective Cost Per Mille, eCPM) that may be obtained by each thousand presentations of the candidate operand. And the benefit score may be calculated according to the following formula:

eCPM＝Pbid*eCTR*eCVR*1000；

wherein eCPM represents a benefit score, pbid represents a bid, eCTR represents a predicted click rate, i.e. the probability of clicking after exposing the operation object to the user, and eCPR represents a predicted conversion rate, i.e. the conversion rate corresponding to the operation object, typically the ratio of the total conversion number of the operation object to the total click number of the operation object is calculated.

After obtaining the profit score, the target weight value can be used for carrying out weighted calculation on the profit score corresponding to the candidate operation object, so as to obtain the target score corresponding to the candidate operation object.

As described above, the target weight value includes the first weight value, the second weight value, and the third weight value, and then the first weight value, the second weight value, and the third weight value may be multiplied by the benefit score to obtain the target score. For example, the target score may be calculated with reference to the following formula;

target score = X Y weight _t *eCPM；

Wherein X is used for representing candidate operation objectsA first weight value, Y is used for representing a second weight value of the candidate operation object, weight _t And a third weight value for representing the candidate operation object, eCPM representing a benefit score of the candidate operation object.

In some embodiments, the candidate operation object may have more than one sub-item attribute, and the foregoing merely ensures that the candidate operation object corresponds to one type of sub-item attribute, and then the candidate operation object may have multiple sub-item attributes of one type, that is, the candidate operation object may have more than one target weight value.

If at least two target weight values exist in the candidate operation object, the target scores corresponding to the target weight values can be calculated respectively, and the final target score with the largest target score is taken as the final target score of the candidate operation object. As described above, if one of the target weight values of a candidate operation object is x2, y1 and weight _t 1, another target weight value is x2, y2 and weight _t 2。

Then a target score of 1 x2 x 1 x weight can be obtained _t 1 eCPM, target score 2 is x2 y2 weight _t 2 ecpm, assuming that the target score 1 is greater than the target score 2, the target score 1 may be determined as the target score of the candidate operation object.

In some embodiments, if at least two target weight values exist in the candidate operation object, calculation may be performed based on the first weight value, the second weight value and the third weight value, and one target weight value is selected from the at least two target weight values to be used as the final target weight value.

It should be noted that, the benefit scores corresponding to the same candidate operation object are the same, the target weight value is in direct proportion to the target score, and the target score is greater as the target weight value is greater, so that the product of the first weight, the second weight value and the third weight value can be calculated first, and the target weight value with the largest product is selected as the target weight value for calculating the target score. As described above, if one of the target weight values of a candidate operation object is x2, y1 and weight _t 1, another target weight value is x2, y2 and weight _t 2, calculating x2 x y1 weight _t 1 to obtain a product 1, calculating x2×y2×weight _t 2 to obtain a product 2, if the product 1 is larger than the product 2, taking the target weight values as x2, y1 and weight _t 1, then, when the target score is calculated later, the target weights used are also x2, y1 and weight _t 1。

After the target score corresponding to each candidate operation object is obtained by calculation, the candidate operation objects may be ranked according to the target score, for example, the candidate operation objects may be ranked according to a mode that the target score is from high to low, a number corresponding to each candidate operation object is obtained, the minimum value of the number is determined as the target operation object, and the target operation object is pushed to a target user.

In one embodiment, if the recommended number of target operation objects pushed to the target user is a plurality of, candidate operation objects with numbers smaller than a preset value may be determined as target operation objects, and the target operation objects may be pushed to the target user together.

In one embodiment, if a specific recommended number can be obtained, the recommended number of candidate operation objects may be sequentially determined in order of from smaller number to larger number, and the candidate operation objects may be pushed to the target user as target operation objects.

In some embodiments, candidate objects may not be ordered, and the target object may be determined directly from the target score. For example, if the recommended number of target operation objects pushed to the target user is 1, the candidate operation object having the largest target score may be determined as the target operation object.

For another example, if the recommended number of target operation objects pushed to the target user is plural, candidate operation objects having a target score greater than a preset score may be determined as target operation objects, and the target operation objects may be pushed to the target user together.

For example, if a specific recommended number can be obtained, the recommended number of candidate operation objects may be sequentially determined according to the order of the target scores from the higher score to the lower score, and the candidate operation objects may be pushed to the target user as target operation objects.

The content recommendation scheme provided by the embodiment of the application can be applied to various recommendation scenes. Such as with advertisement recommendations, video recommendations, music recommendations, book recommendations, content recommendations in smart sounds, etc. The method provided by the embodiment of the application can acquire the operation time, the operation data and the operation object of the target user, acquire the object attribute corresponding to the operation object, determine the candidate operation object possibly interested by the target user by using the object attribute, and assign different target weight values to the candidate operation object based on the object attribute and the operation time, thereby fully considering the influence of the user's interest, ensuring that the target operation object recommended to the target user is related to the user's interest, and further improving the recommendation accuracy.

The method described in the above embodiments will be described in further detail below.

In this embodiment, a method according to an embodiment of the present application will be described in detail using advertisement recommendation as an example.

As shown in fig. 2, a specific flow of a content recommendation method is as follows:

s210, acquiring operation time, operation data and advertisements corresponding to the operation data of a target user.

S220, acquiring the object attribute corresponding to the advertisement.

S230, determining candidate advertisements corresponding to the target users according to the object attributes.

S240, calculating a target weight value corresponding to each candidate advertisement based on the operation time and the article attribute.

S250, determining target advertisements from the candidate advertisements based on the target weight values, and pushing the target advertisements to target users.

After the operation time, the operation data and the corresponding operation objects of the target user are acquired, the operation time, the operation data and the advertisements can be considered to have an association relation, namely, when the target user operates at the operation time, a certain operation is performed on the advertisements, and the certain operation can be acquired through the operation data. For example, the target user clicks on advertisement 1 at 12 points, and the target user installs game application 2 via advertisement 2 at 13 points.

Assuming that advertisement 1 is an advertisement of a Z-brand mobile phone and advertisement 2 is an advertisement of application 2, then, for advertisement 1, the attribute of the item corresponding to the advertisement may be obtained, where the identifier of the advertisement is advertisement 1, the first class category of the item corresponding to advertisement 1 is a digital product, the second class category of the item corresponding to advertisement 1 is a mobile phone, the third class category corresponding to advertisement 1 is a Z-brand, that is, attribute 1 of the item corresponding to advertisement 1 may be: advertisement 1-digital product-cell phone-Z brand, similarly, the item attribute 2 corresponding to advertisement 2 may be: advertisement 2-software product-application-game.

After the item attribute is acquired, candidate advertisements corresponding to the target user can be determined according to the item attribute 1 and the item attribute 2.

For example, advertisement queues corresponding to each child item attribute are pre-bound, and when candidate advertisements corresponding to the target user are determined, the corresponding advertisement queues can be pulled based on the item attribute. Therefore, the advertisement queues corresponding to the advertisement 1, the advertisement queues corresponding to the digital product, the advertisement queues corresponding to the mobile phone, the advertisement queues corresponding to the Z brand, the advertisement queues corresponding to the advertisement 2, the advertisement queues corresponding to the software product, the advertisement queues corresponding to the application program and the advertisement queues corresponding to the game can be obtained, and the advertisement queues are 8 advertisement queues in total, and the advertisements in the 8 advertisement queues are candidate advertisements.

After the candidate advertisements are pulled, a target weight value for each candidate advertisement may be calculated, where the target weight value includes a first weight value, a second weight value, and a third weight value. First, since 8 advertisement queues are pulled, there may be repeated advertisements in the advertisement queues, so that repeated advertisements may be merged, for example, if advertisement 5 is in the advertisement queue corresponding to the software product and the advertisement queue corresponding to the Z brand at the same time, since the Z brand is under the three-level category, advertisement 5 may preserve the sub-item attribute of the Z brand.

For another example, the advertisement 6 is in the advertisement queue corresponding to the software product and the advertisement queue corresponding to the mobile phone, and because the two identifiers are in the second class, the two sub-item attributes can be reserved for the advertisement 6 at the same time.

In calculating the target weight value of the advertisement 5, it may be that the child item attribute and the operation time used for pulling the advertisement 5 are determined: z brand-12 points, since Z brand is the sign of click advertisement 1, the corresponding time is 12 points. Then, the first weight value can be determined according to the Z brand, the second weight value can be determined according to clicking, the third weight value can be determined according to 12 points, and the target weight value 1 corresponding to the advertisement 5 can be obtained.

When the target weight value of the advertisement 5 is calculated, two sub-object attributes corresponding to the advertisement 5 are pulled, namely the mobile phone is 12 points respectively, the corresponding time is 12 points due to the fact that the identifier of the mobile phone is generated by clicking the advertisement 1, the corresponding time is 13 points due to the fact that the identifier of the software product is generated by downloading an application program through the advertisement 2, and the corresponding time is 13 points. Thus, advertisement 5 may have two target weight values, e.g., cell phone-12 points corresponding to target weight value 1 and software product-13 points corresponding to target weight value 2.

In general, when calculating the scores of advertisements and ranking the advertisements, the scores are mainly based on eCPM, however, advertisements with low eCPM may be interested by the user, but advertisements with low eCPM cannot be recommended to the user, and the recommended advertisements are not interested by the user easily, so that the accuracy of the recommendation is reduced, if the advertisements are recommended only depending on the interests of the user, the bid and the waste of the traffic value of the advertisements may be possibly caused, therefore, in order to consider the traffic value of the advertisements and the interests of the user, the eCPM of the advertisements may be used as a base score, the interests of the users may be converted into target weight values, and the target scores of the candidate advertisements may be calculated according to the target weight values and the eCPM scores of the advertisements, so as to determine the first N candidate advertisements with the target scores being the target advertisements.

The target score corresponding to the advertisement 5 may be calculated by multiplying the eCPM score of the advertisement 5 by the corresponding target weight value.

Since two target weight values exist in the advertisement 6, the corresponding target score under the target weight value 1 may be that the target weight value 1 is multiplied by the eCPM score of the advertisement 6, the corresponding target score under the target weight value 2 may be that the target weight value 2 is multiplied by the eCPM score of the advertisement 6, and then the target score is determined to be the target score of the advertisement 6.

Thus, the target scores of all candidate advertisements can be calculated, and the target advertisements are sequentially pushed to target users according to the sequence of the target scores from high to low.

It can be seen from the foregoing that, according to the content recommendation method provided by the embodiment of the present application, not only the operation object actually operated by the user is considered, but also the object indicated by the operation object is considered, the object attribute is generated, the candidate operation object which may be interested by the user is determined based on the object attribute, the influence of the actual operation of the user is considered, the target weight value is calculated, the eCPM score is influenced by using the target weight value, the eCPM score and the influence of the actual operation of the user are considered, and the recommendation which better meets the user interest is realized, thereby improving the recommendation accuracy.

In order to better implement the method, the embodiment of the application also provides a content recommendation device, which can be integrated in an electronic device, wherein the electronic device can be a terminal, a server and other devices. The terminal can be a mobile phone, a tablet computer, an intelligent Bluetooth device, an intelligent voice interaction device, an intelligent household appliance, a vehicle-mounted terminal, an aircraft, a notebook computer or a personal computer, or a server cluster formed by a plurality of servers.

For example, in the present embodiment, a method according to an embodiment of the present application will be described in detail by taking a specific integration of a content recommendation device in a server as an example.

For example, as shown in fig. 3, the content recommendation device 300 may include a first acquisition module 310, a second acquisition module 320, a candidate determination module 330, a weight calculation module 340, and a recommendation module 350.

A first obtaining module 310, configured to obtain an operation time of a target user, operation data, and an operation object corresponding to the operation data;

a second obtaining module 320, configured to obtain an item attribute corresponding to the operation object;

a candidate determining module 330, configured to determine a candidate operation object corresponding to the target user according to the item attribute;

a weight calculation module 340, configured to calculate a target weight value of each candidate operation object according to the operation time and the item attribute;

and a recommendation module 350, configured to determine, based on the target weight value, a target operation object recommended to the target user from the candidate operation objects.

In some embodiments, the recommendation module 350 further comprises:

In some embodiments, the item attribute includes a plurality of sub-item attributes, the target weight value includes a first weight value, a second weight value, and a third weight value, and the weight calculation module 340 further includes:

In some embodiments, the item attributes include a plurality of sub-item attributes, and the candidate determination module 330 further includes:

In some embodiments, the sub-candidate determination unit is further to:

In the implementation, each module or unit may be implemented as an independent entity, or may be implemented as the same entity or several entities in any combination, and the implementation of each module or unit may be referred to the foregoing method embodiments and will not be repeated herein.

As can be seen from the above, the content recommendation device of the present embodiment may obtain the operation time, the operation data and the operation object of the target user, obtain the item attribute corresponding to the operation object, determine the candidate operation object that may be interested by the target user using the item attribute, and assign different target weight values to the candidate operation object based on the item attribute and the operation time, so as to fully consider the influence of the user's interest, and ensure that the target operation object recommended to the target user is related to the user's interest, thereby improving the recommendation accuracy.

The embodiment of the application also provides electronic equipment which can be a terminal, a server and other equipment. The terminal can be a mobile phone, a tablet computer, an intelligent Bluetooth device, a notebook computer, a personal computer and the like; the server may be a single server, a server cluster composed of a plurality of servers, or the like.

In this embodiment, a detailed description will be given taking an example that the electronic device of this embodiment is a server, for example, as shown in fig. 4, which shows a schematic structural diagram of the server according to the embodiment of the present application, specifically:

The server may include one or more processors 401 of a processing core, memory 402 of one or more computer readable storage media, a power supply 403, an input module 404, and a communication module 405, among other components. Those skilled in the art will appreciate that the server architecture shown in fig. 4 is not limiting of the server and may include more or fewer components than shown, or may combine certain components, or a different arrangement of components. Wherein:

the processor 401 is a control center of the server, connects respective portions of the entire server using various interfaces and lines, and performs various functions of the server and processes data by running or executing software programs and/or modules stored in the memory 402, and calling data stored in the memory 402. In some embodiments, processor 401 may include one or more processing cores; in some embodiments, processor 401 may integrate an application processor that primarily processes operating systems, user interfaces, applications, and the like, with a modem processor that primarily processes wireless communications. It will be appreciated that the modem processor described above may not be integrated into the processor 401.

The memory 402 may be used to store software programs and modules, and the processor 401 executes various functional applications and data processing by executing the software programs and modules stored in the memory 402. The memory 402 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like; the storage data area may store data created according to the use of the server, etc. In addition, memory 402 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 with access to the memory 402.

The server also includes a power supply 403 for powering the various components, and in some embodiments, the power supply 403 may be logically connected to the processor 401 by a power management system, such that charge, discharge, and power consumption management functions are performed by the power management system. The power supply 403 may also include one or more of any of a direct current or alternating current power supply, a recharging system, a power failure detection circuit, a power converter or inverter, a power status indicator, and the like.

The server may also include an input module 404, which input module 404 may be used to receive entered numeric or character information and to generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

The server may also include a communication module 405, and in some embodiments the communication module 405 may include a wireless module, through which the server may wirelessly transmit over short distances, thereby providing wireless broadband internet access to the user. For example, the communication module 405 may be used to assist a user in e-mail, browsing web pages, accessing streaming media, and so forth.

Although not shown, the server may further include a display unit or the like, which is not described herein. In this embodiment, the processor 401 in the server loads executable files corresponding to the processes of one or more application programs into the memory 402 according to the following instructions, and the processor 401 executes the application programs stored in the memory 402, so as to implement various functions as follows:

Acquiring an object attribute corresponding to the operation object;

The specific implementation of each operation above may be referred to the previous embodiments, and will not be described herein.

As can be seen from the above, the embodiment of the present application may acquire the operation time, the operation data and the operation object of the target user, acquire the item attribute corresponding to the operation object, determine the candidate operation object that may be of interest to the target user by using the item attribute, and assign different target weight values to the candidate operation object based on the item attribute and the operation time, so as to fully consider the influence of the interest of the user, and ensure that the target operation object recommended to the target user is related to the interest of the user, thereby improving the recommendation accuracy.

Those of ordinary skill in the art will appreciate that all or a portion of the steps of the various methods of the above embodiments may be performed by instructions, or by instructions controlling associated hardware, which may be stored in a computer-readable storage medium and loaded and executed by a processor.

To this end, an embodiment of the present application provides a computer readable storage medium having stored therein a plurality of instructions capable of being loaded by a processor to perform the steps of any of the content recommendation methods provided by the embodiments of the present application. For example, the instructions may perform the steps of:

acquiring an object attribute corresponding to the operation object;

Wherein the storage medium may include: read Only Memory (ROM), random access Memory (RAM, random Access Memory), magnetic or optical disk, and the like.

According to one aspect of the present application, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The computer instructions are read from a computer-readable storage medium by a processor of a computer device, and executed by the processor, cause the computer device to perform the methods provided in various alternative implementations of the content recommendation aspects provided in the above-described embodiments.

The instructions stored in the storage medium may perform steps in any content recommendation method provided by the embodiments of the present application, so that the beneficial effects that any content recommendation method provided by the embodiments of the present application can be achieved are detailed in the previous embodiments, and are not repeated here.

The foregoing has described in detail a content recommendation method, apparatus, electronic device, storage medium and program product provided by the embodiments of the present application, and specific examples are applied herein to illustrate the principles and embodiments of the present application, and the above description of the embodiments is only for aiding in understanding the method and core idea of the present application; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in light of the ideas of the present application, the present description should not be construed as limiting the present application.

Claims

1. A content recommendation method, the method comprising:

acquiring an object attribute corresponding to the operation object;

2. The method of claim 1, wherein the determining, based on the target weight value, a target operation object recommended to the target user from the candidate operation objects, comprises:

obtaining a benefit score corresponding to the candidate operation object;

weighting calculation is carried out on the benefit score based on the target weight value, so that a target score is obtained;

and determining a target operation object recommended to the target user from candidate operation objects based on the target score.

3. The method of claim 1, wherein the item attributes comprise a plurality of sub-item attributes, the target weight values comprise a first weight value, a second weight value, and a third weight value, and the calculating the target weight value for each candidate operation object based on the operation time and the item attributes comprises:

Determining the first weight value according to the attribute of the sub-item used by the determined candidate operation object;

determining the second weight value of the candidate operation object according to the operation data corresponding to the child object attribute;

and calculating the third weight value according to the difference value between the operation time and the current time.

4. The method of claim 1, wherein the item attributes comprise a plurality of sub-item attributes, and wherein the determining the candidate operation object corresponding to the target user according to the item attributes comprises:

determining a sub candidate operation object corresponding to each sub item attribute according to each sub item attribute;

and determining all the sub candidate operation objects as the candidate operation objects.

5. The method of claim 4, wherein determining a child candidate operation object corresponding to each child item attribute according to the each child item attribute comprises:

6. The method of claim 4, wherein determining a child candidate operation object corresponding to each child item attribute according to the each child item attribute comprises:

7. A content recommendation device, the device comprising:

8. An electronic device comprising a processor and a memory, the memory storing a plurality of instructions; the processor loads instructions from the memory to perform the steps in the content recommendation method according to any of claims 1 to 6.

9. A computer readable storage medium storing a plurality of instructions adapted to be loaded by a processor to perform the steps of the content recommendation method of any one of claims 1 to 6.

10. A computer program product comprising computer programs/instructions which when executed by a processor implement the steps of the content recommendation method according to any one of claims 1 to 6.