CN114201677A

CN114201677A - Material recommendation method and device, electronic equipment, storage medium and product

Info

Publication number: CN114201677A
Application number: CN202111503473.XA
Authority: CN
Inventors: 蒋俊翔; 徐伟; 夏晓玲; 曹丞泰; 何伯磊; 陈坤斌; 和为
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2021-12-09
Filing date: 2021-12-09
Publication date: 2022-03-18

Abstract

The disclosure provides a material recommendation method, a material recommendation device, electronic equipment, a storage medium and a product, and relates to the technical field of artificial intelligence, in particular to the technical field of big data and cloud computing. The specific implementation scheme is as follows: acquiring an account vector corresponding to a target account in an account vector library, wherein the account vector is determined based on the incidence relation between the account and reference data; based on the account vector, obtaining a material vector associated with the account vector in a material vector library, wherein the material vector is determined based on the association relationship between materials and accounts; and recommending the materials corresponding to the material vectors to the target account. The accuracy of recommending materials can be improved through the method and the device.

Description

Material recommendation method and device, electronic equipment, storage medium and product

Technical Field

The disclosure relates to the technical field of artificial intelligence, in particular to the technical field of big data and cloud computing, and specifically relates to a material recommendation method and device, electronic equipment, a storage medium and a product.

Background

In an office scenario, the required materials can be recommended for the account in general. When recommending materials for an account, materials related to the interest of the account can be recommended according to the interest of the account. That is, the recommended material needs to have information related to interest.

Disclosure of Invention

The disclosure provides a material recommendation method, a material recommendation device, electronic equipment, a storage medium and a product.

According to a first aspect of the present disclosure, there is provided a material recommendation method, the method comprising:

acquiring an account vector corresponding to a target account in an account vector library, wherein the account vector is determined based on the incidence relation between the account and reference data; based on the account vector, obtaining a material vector associated with the account vector in a material vector library, wherein the material vector is determined based on the association relationship between materials and accounts; and recommending the materials corresponding to the material vectors to the target account.

According to a second aspect of the present disclosure, there is provided a material recommendation device, the device comprising:

the acquisition module is used for acquiring an account vector corresponding to the target account from an account vector library, wherein the account vector is determined based on the incidence relation between the account and the reference data; the system is also used for acquiring a material vector related to the account vector in a material vector library based on the account vector, wherein the material vector is determined based on the incidence relation between materials and accounts; and the recommending module is used for recommending the materials corresponding to the material vectors to the target account.

According to a third aspect of the present disclosure, there is provided an electronic device comprising:

at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of the first aspect.

According to a fourth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method according to the first aspect.

According to a fifth aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the method according to the first aspect.

It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.

Drawings

The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

fig. 1 is a schematic flow chart illustrating a material recommendation method provided by an embodiment of the present disclosure;

FIG. 2 is a block diagram illustrating a method for recommending materials according to an embodiment of the present disclosure;

fig. 3 is a schematic flowchart illustrating a process of determining an account vector in a material recommendation method according to an embodiment of the present disclosure;

fig. 4 is a schematic flowchart illustrating a process of determining an account vector in a material recommendation method according to an embodiment of the present disclosure;

fig. 5 is a schematic flow chart illustrating a determination of a skill tag prediction model in a material recommendation method according to an embodiment of the present disclosure;

fig. 6 shows a schematic structural diagram of a skill tag prediction model in a material recommendation method provided by an embodiment of the present disclosure;

fig. 7 is a schematic flow chart illustrating determining a material vector in a material recommendation method according to an embodiment of the present disclosure;

fig. 8 is a schematic flow chart illustrating determining a material vector in a material recommendation method according to an embodiment of the present disclosure;

fig. 9 is a schematic structural diagram illustrating vector determination in a material recommendation method according to an embodiment of the present disclosure;

fig. 10 is a schematic flow chart illustrating determining a material vector in a material recommendation method according to an embodiment of the present disclosure;

fig. 11 is a schematic structural diagram illustrating a material recommending apparatus provided by an embodiment of the present disclosure;

FIG. 12 is a block diagram of an electronic device for implementing a method of material recommendation of an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

In an office scenario, materials related to the work content of an account at work can be recommended. In a recommendation system, a recommendation method is generally adopted to recommend materials of interest to a user using an account through clicked materials based on click behavior. In other words, the clicking behavior of the user reflects the interest of the user, and the materials are recommended for the user according to the interest of the user.

However, when the click behavior is less or the click behavior is not related to the work, the recommended materials for the user cannot be applied to the office scene of the user. Therefore, materials recommended by using the material recommendation method in the related art in an office scene cannot meet the working requirements of users.

Based on the defects of the related art in the above embodiments, the present disclosure provides a material recommendation method, which can recommend a material related to work for an account in an office scene by using reference data of the account related to a user, and the recommended material better conforms to the work skill of the user using the account.

The following embodiments will explain the material recommending method of the present disclosure with reference to the accompanying drawings.

Fig. 1 shows a schematic flowchart of a material recommendation method provided in an embodiment of the present disclosure, and as shown in fig. 1, the method may include:

in step S110, an account vector corresponding to the target account is acquired in the account vector library.

In embodiments of the present disclosure, the account vector is determined based on an association of the account with the reference data. The reference data may be data related to work information, for example, the reference data may be completed work information, unfinished work information, set work targets, and the like.

In the present disclosure, a skill label for an account may be derived based on reference data for the account. And encoding the skill tags of the account and the skill tags of the reference data to obtain an account vector after encoding, and storing the account vector in an account vector library. Note that the account vector may be predetermined.

In step S120, a material vector associated with the account vector is obtained in a material vector library based on the account vector.

In embodiments of the present disclosure, the item vector is determined based on the association of the item with the account.

In the present disclosure, the material vector may be determined in a similar manner to the account vector, and the skill tags of the material and the account may be encoded to determine the material vector. And storing the obtained material vector into a material vector library.

At least one material vector associated with the account is obtained by calculating the account vector and the material vectors in the material vector library.

In step S130, materials corresponding to the material vector are recommended to the target account.

In the embodiment of the disclosure, according to the obtained at least one material vector, determining a material corresponding to each material vector, and recommending the determined at least one material to the target account.

Fig. 2 is a schematic diagram illustrating a framework of a material recommending method according to an embodiment of the present disclosure, and as shown in fig. 2, the material recommending method may be divided into an upper line part and a lower line part. The offline part can be completed in advance, namely, the skill label of the account and the skill label of the material are determined, the account vector is obtained through model coding according to the incidence relation between the account and the reference data, and the obtained account vector is stored in an account vector library. And obtaining a material vector through model coding according to the incidence relation between the material and the account, and storing the obtained material vector into a material vector library. The online part performs operations, after the target account is determined, according to the identification or other information of the account, the account vector corresponding to the target account is acquired in the account vector library. And performing vector calculation according to the determined account vector and the material vectors in the material vector library to determine at least one material, such as material 1, material 2 and the like.

Therefore, according to the material recommending method provided by the embodiment of the disclosure, the materials related to the account and the reference data can be determined according to the association relationship between the reference data of the account and the association relationship between the account and the materials, so that the accuracy of recommending the materials for the account is improved.

As described above, in the present disclosure, the account vector may be predetermined. The following embodiments will describe determining an account vector based on an association of an account with reference data.

Fig. 3 shows a schematic flowchart of determining an account vector in a material recommendation method provided by an embodiment of the present disclosure, and as shown in fig. 3, the method may include:

in step S210, reference data of the account is obtained, and the reference data is input to the skill label prediction model to obtain a first skill label of the account.

In the embodiment of the present disclosure, the reference data may be work information used by the account in an office scenario, for example, content of work obtained through the work information, incomplete work, and the like, and may also be other work data, which is not specifically limited in the present disclosure. It should be noted that all the work information is obtained under the condition that the user agrees and allows.

In the method, reference data of an account related to a user are acquired, the reference data are input into a skill label prediction model obtained through pre-training, and the skill label prediction model outputs a first skill label of the account.

In step S220, sample data of the reference data is acquired.

In the embodiment of the present disclosure, sample data of the reference data also needs to be acquired, and the sample data may be historical reference data. Wherein the sample data comprises positive sample reference data associated with the account and negative sample reference data not associated with the account. In other words, reference data acquired based on an account is taken as positive exemplar reference data, and reference data other than the account is taken as negative exemplar reference data.

In step S230, the first skill label of the positive sample reference data and the first skill label of the negative sample reference data are expressed by using the dense vector, resulting in an account vector.

In the present disclosure, after determining positive and negative exemplar reference data of sample data, a first skill label possessed by the positive exemplar reference data and a first skill label possessed by the negative exemplar reference data are acquired.

And coding the first skill label of the account according to the correlation between the first skill label of the positive sample reference data and the first skill label of the account and the correlation between the first skill label of the negative sample reference data and the first skill label of the account to obtain an account vector.

In the embodiment of the disclosure, the skill tag in the reference data is used as an auxiliary tag of the account skill tag, and the account skill tag is supplemented, so that the obtained account vector can more comprehensively characterize the characteristics of the account.

The following example will further illustrate the derivation of an account vector.

Fig. 4 is a schematic flowchart illustrating a process of determining an account vector in a material recommendation method provided by an embodiment of the present disclosure, and as shown in fig. 4, the method may include:

in step S310, a probability ranking of the first skill tag is obtained based on the skill tag prediction model.

In the embodiment of the disclosure, the first skill tag output by the skill tag prediction model further comprises a probability value corresponding to the first skill tag. That is, if there are a plurality of first skill tags, the skill tag prediction model may also give a probability value of each first skill tag. Thereby determining the probability that the dealer owns the label. And sequencing the obtained first skill tags according to the probability value of each first skill tag to obtain the probability sequencing of the first skill tags.

In step S320, the first skill label of the positive exemplar reference data and the first skill label of the negative exemplar reference data are encoded based on the probability ordering.

In embodiments of the present disclosure, a first skill tag of positive exemplar reference data and a first skill tag of negative exemplar reference data are encoded according to a probability that an account has a skill tag.

In the present disclosure, the first skill tag may be tagged using a Bidirectional Encoder Representation (BERT).

In step S330, vector mapping is performed on the encoded first skill label of the positive sample reference data and the encoded first skill label of the negative sample reference data to obtain an account vector.

In the embodiment of the present disclosure, the first skill label of the encoded positive sample reference data and the first skill label of the negative sample reference data are used as inputs of a Multilayer Perceptron (MLP), and the MLP may perform vector mapping on a relationship between the skill labels and accounts, and output account vectors, that is, obtain the account vectors.

By using MLP in the disclosure, the expression capability of the vector can be improved, and the function of reducing the dimension of the skill label is achieved.

As in the above example, the skill tag prediction model is predetermined, and the implementation of determining the skill tag prediction model is as follows.

Fig. 5 is a schematic flow chart illustrating the determination of the skill tag prediction model in the material recommendation method provided by the embodiment of the disclosure, as shown in fig. 5, including the following steps.

In step S410, keywords corresponding to the skill tags are extracted from the reference data, and the correspondence between the keywords and the skill tags is obtained.

In the embodiment of the present disclosure, in the reference data, keywords corresponding to skill tags are extracted. For example, in the work information of the account, a keyword corresponding to the skill tag is extracted, and the correspondence between the keyword and the skill tag is obtained, that is, a skill tag-keyword pair is formed.

In step S420, historical reference data of the account is obtained, and a historical first skill tag of the account is determined according to the corresponding relationship.

In the disclosed embodiment, historical first skill tags are marked for all accounts according to the formed skill tag-keyword pairs and historical reference data of the accounts.

In step S430, the skill label prediction model is obtained by using the historical reference data as input and the historical first skill label as an output training model.

In the embodiment of the disclosure, historical reference data is used as input, first skill labels of all accounts are used as output, a supervision model is trained, and the generalization ability of the accounts is supplemented. And iteratively training the supervision model until the skill label of the account can be accurately predicted by the output of the trained supervision model, and stopping training to obtain a skill label prediction model. Wherein the account has a skill tag characterizing the use of the user's skill tag.

Fig. 6 shows a schematic structural diagram of a skill tag prediction model in the material recommendation method provided by the embodiment of the disclosure, and as shown in fig. 6, according to the work information and the set rule, the keyword corresponding to the skill tag is extracted, and the skill tag data is determined according to the corresponding relationship between the skill tag and the keyword. Training according to the working information and the skill label data, and continuously iterating the process until a skill label prediction model capable of accurately predicting the skill label is obtained.

In the disclosure, the skill label prediction model obtained by obtaining the skill label-keyword pair and further using the skill label-keyword pair training model has better generalization capability.

Fig. 7 shows a schematic flowchart of determining a material vector in a material recommendation method provided by an embodiment of the present disclosure, and as shown in fig. 7, the method may include:

in step S510, a material is obtained, and a second skill tag of the material is determined according to a corresponding relationship between the keyword and the skill tag.

In the embodiment of the disclosure, the existing material is obtained, and the corresponding second skill tag is determined for the existing material according to the determined corresponding relationship between the keyword and the skill tag.

In step S520, a first skill tag of the account is obtained, and according to the association relationship between the first skill tag and the second skill tag, it is determined that the material related to the account is a positive sample material, and the material unrelated to the account is a negative sample material.

In the embodiment of the disclosure, according to the obtained first skill tag, in the second skill tag, determining the association relationship between the first skill tag and the second skill tag. For example, the association between the first skill tag and the second skill tag is obtained from the same skill tag of the first skill tag and the second skill tag. And determining the materials corresponding to the second skill tags intersected with the first skill tags as the materials related to the account as positive sample materials. And determining the materials corresponding to the second skill tags without intersection with the first skill tags as the materials irrelevant to the account as negative sample materials.

In step S530, the second skill labels of the positive sample materials and the second skill labels of the negative sample materials are represented by dense vectors, so as to obtain material vectors.

And expressing the skill labels of the materials by using dense vectors according to the correlation between the second skill label and the first skill label of the positive sample material and the correlation between the second skill label and the first skill label of the negative sample material to obtain a material vector.

In the disclosure, based on the correlation between physics and an account, the skill labels of materials are represented by using dense vectors, and the obtained material vectors can enhance the expression capability of the correlation between the account and the materials.

Fig. 8 is a schematic flowchart illustrating a process of determining a material vector in a material recommendation method provided by an embodiment of the present disclosure, and as shown in fig. 8, the method may include:

in step S610, a probability ranking of the first skill tags is obtained, and the second skill tags are ranked based on the probability ranking, so as to obtain a probability ranking of the second skill tags.

In the embodiment of the present disclosure, as described above, the probability value corresponding to the first skill tag output by the skill tag prediction model is obtained, the probability value of the second skill tag corresponding to the first skill tag is determined according to the probability value of each first skill tag, and the second skill tags are ranked based on the probability values, so as to obtain the probability ranking of the second skill tags.

In step S620, the second skill tags of positive sample material and the second skill tags of negative sample material are encoded based on the probability ranking.

In embodiments of the present disclosure, the second skill tags for positive sample material and the second skill tags for negative sample material are encoded according to a probabilistic ranking of the second skill tags.

In the present disclosure, the first skill tag may also be tagged using BERT.

In step S630, vector mapping is performed on the encoded second skill label of the positive sample material and the encoded second skill label of the negative sample material to obtain a material vector.

In the embodiment of the disclosure, the second skill label of the encoded positive sample material and the second skill label of the negative sample material are used as the input of the MLP, and the MLP may perform vector mapping on the relationship between the skill labels and the account, and output the material vector, so as to obtain the material vector.

Fig. 9 shows a structural diagram of vector determination in a material recommendation method provided by an embodiment of the present disclosure, as shown in fig. 9, including a skill tag of reference data, a skill tag of an account, and a skill tag of a material. And respectively encoding the skill tags of the reference data, the account and the materials by using BERT, and inputting the encoded skill tags of the reference data, the account and the materials into the MLP to obtain an account vector and a material vector. In the figure, task 1 is auxiliary information, and task 2 is an account vector and a material vector which are required to be obtained by the present disclosure.

Fig. 10 is a schematic flowchart illustrating a process of determining a material vector in a material recommendation method provided by an embodiment of the present disclosure, and as shown in fig. 10, the method may include:

in step S710, an account vector and a material vector are calculated to obtain a vector calculation result.

In step S720, a material vector associated with the account vector is determined in the material vector library according to the vector calculation result.

In the embodiment of the disclosure, the account vector and the material vector can be calculated in the material vector library through anns-serving to obtain a vector calculation result. Wherein the vector calculation result can be that the materials and the account have the same skill label. Thus, in the material vector library, the material related to the account is determined.

In the embodiment of the disclosure, in an office scene, the materials recommended for the account by the material recommendation method can be recommended for the account under the office scene, and materials which can be used for reference and correct guidance can be recommended for the account, so that the working efficiency of a user is improved.

Based on the same principle as the method shown in fig. 1, fig. 11 shows a schematic structural diagram of a material recommending apparatus provided by an embodiment of the disclosure, and as shown in fig. 11, the material recommending apparatus 100 may include:

an obtaining module 101, configured to obtain an account vector corresponding to a target account in an account vector library, where the account vector is determined based on an association relationship between an account and reference data; the system is also used for acquiring a material vector related to the account vector in a material vector library based on the account vector, wherein the material vector is determined based on the incidence relation between materials and accounts; and the recommending module 102 is configured to recommend the materials corresponding to the material vector to the target account.

In the embodiment of the present disclosure, the material recommending apparatus further includes a determining module 103.

The determining module 103 is configured to obtain reference data of an account, input the reference data into a skill tag prediction model, and obtain a first skill tag of the account; obtaining sample data of reference data, the sample data comprising positive sample reference data related to the account and negative sample reference data not related to the account; and expressing the first skill label of the positive sample reference data and the first skill label of the negative sample reference data by using a dense vector to obtain an account vector.

A determining module 103, configured to obtain a probability ranking of the first skill tag based on the skill tag prediction model; encoding a first skill label of the positive exemplar reference data and a first skill label of the negative exemplar reference data based on the probability ordering; and carrying out vector mapping on the encoded first skill label of the positive sample reference data and the encoded first skill label of the negative sample reference data to obtain an account vector.

The determining module 103 is configured to extract a keyword corresponding to the skill tag from the reference data, and obtain a corresponding relationship between the keyword and the skill tag; acquiring historical reference data, and determining a historical first skill tag of the historical reference data according to the corresponding relation; and taking the historical reference data as input, and taking the historical first skill label as an output training model to obtain a skill label prediction model.

The determining module 103 is used for acquiring a material and determining a second skill tag of the material according to the corresponding relation between the keyword and the skill tag; acquiring a first skill label of an account, and determining that materials related to the account are positive sample materials and materials unrelated to the account are negative sample materials according to the incidence relation between the first skill label and a second skill label; and expressing the second skill label of the positive sample material and the second skill label of the negative sample material by using a dense vector to obtain a material vector.

The determining module 103 is configured to obtain a probability ranking of the first skill label, and rank the second skill label based on the probability ranking to obtain a probability ranking of the second skill label. Coding a second skill label of the positive sample material and a second skill label of the negative sample material based on the probability ordering; and carrying out vector mapping on the encoded second skill label of the positive sample material and the encoded second skill label of the negative sample material to obtain a material vector.

The obtaining module 101 is configured to calculate an account vector and a material vector to obtain a vector calculation result. And determining a material vector associated with the account vector in the material vector library according to the vector calculation result.

In the technical scheme of the disclosure, the acquisition, storage, application and the like of the personal information of the related user all accord with the regulations of related laws and regulations, and do not violate the good customs of the public order.

The present disclosure also provides an electronic device, a readable storage medium, and a computer program product according to embodiments of the present disclosure.

FIG. 12 shows a schematic block diagram of an example electronic device 200 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 12, the apparatus 200 includes a computing unit 201 that can perform various appropriate actions and processes in accordance with a computer program stored in a Read Only Memory (ROM)202 or a computer program loaded from a storage unit 208 into a Random Access Memory (RAM) 203. In the RAM 203, various programs and data required for the operation of the device 200 can also be stored. The computing unit 201, the ROM 202, and the RAM 203 are connected to each other via a bus 204. An input/output (I/O) interface 205 is also connected to bus 204.

Various components in the device 200 are connected to the I/O interface 205, including: an input unit 206 such as a keyboard, a mouse, or the like; an output unit 207 such as various types of displays, speakers, and the like; a storage unit 208, such as a magnetic disk, optical disk, or the like; and a communication unit 209 such as a network card, modem, wireless communication transceiver, etc. The communication unit 209 allows the device 200 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The computing unit 201 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 201 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various dedicated Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The computing unit 201 performs the various methods and processes described above, such as method material recommendations. For example, in some embodiments, the method material recommendation may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 208. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 200 via the ROM 202 and/or the communication unit 209. When loaded into RAM 203 and executed by computing unit 201, may perform one or more of the steps of the method material recommendation described above. Alternatively, in other embodiments, the computing unit 201 may be configured to perform the method material recommendation in any other suitable manner (e.g., by way of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server with a combined blockchain.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel or sequentially or in different orders, and are not limited herein as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved.

The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A method of material recommendation, the method comprising:

acquiring an account vector corresponding to a target account in an account vector library, wherein the account vector is determined based on the incidence relation between the account and reference data;

based on the account vector, obtaining a material vector associated with the account vector in a material vector library, wherein the material vector is determined based on the association relationship between materials and accounts;

and recommending the materials corresponding to the material vectors to the target account.

2. The method of claim 1, wherein the determining an account vector based on the association of the account with the reference data comprises:

acquiring reference data of an account, and inputting the reference data into a skill tag prediction model to obtain a first skill tag of the account;

obtaining sample data of reference data, the sample data comprising positive sample reference data related to the account and negative sample reference data not related to the account;

and representing the first skill label of the positive sample reference data and the first skill label of the negative sample reference data by using a dense vector to obtain an account vector.

3. The method of claim 2, wherein the representing the first skill label of the positive exemplar reference data and the first skill label of the negative exemplar reference data using a dense vector, resulting in an account vector, comprises:

obtaining a probability ranking of the first skill label based on the skill label prediction model;

encoding a first skill label of the positive exemplar reference data and a first skill label of the negative exemplar reference data based on the probability ordering;

and carrying out vector mapping on the encoded first skill label of the positive sample reference data and the encoded first skill label of the negative sample reference data to obtain an account vector.

4. The method of claim 2, wherein determining the skill tag prediction model comprises:

extracting keywords corresponding to the skill tags from the reference data to obtain corresponding relations between the keywords and the skill tags;

acquiring historical reference data, and determining a historical first skill tag of the historical reference data according to the corresponding relation;

and taking the historical reference data as input, and taking the historical first skill label as an output training model to obtain a skill label prediction model.

5. The method of claim 1, wherein the determining a material vector based on the association of materials to accounts comprises:

obtaining a material, and determining a second skill label of the material according to the corresponding relation between the keyword and the skill label;

acquiring a first skill label of an account, and determining that materials related to the account are positive sample materials and materials unrelated to the account are negative sample materials according to the incidence relation between the first skill label and a second skill label;

and expressing the second skill label of the positive sample material and the second skill label of the negative sample material by using a dense vector to obtain a material vector.

6. The method of claim 5, wherein deriving a material vector based on representing the second skill label of the positive sample material and the second skill label of the negative sample material using a dense vector comprises:

acquiring the probability sequence of the first skill label, and sequencing a second skill label based on the probability sequence to obtain the probability sequence of the second skill label;

encoding a second skill label for the positive sample material and a second skill label for the negative sample material based on the probability rankings;

and carrying out vector mapping on the encoded second skill label of the positive sample material and the encoded second skill label of the negative sample material to obtain a material vector.

7. The method of claim 1, wherein the obtaining, based on the account vector, a material vector associated with the account vector in a material vector repository comprises:

calculating the account vector and the material vector to obtain a vector calculation result;

and determining a material vector associated with the account vector in a material vector library according to the vector calculation result.

8. A material recommendation device, the device comprising:

the acquisition module is used for acquiring an account vector corresponding to the target account from an account vector library, wherein the account vector is determined based on the incidence relation between the account and the reference data; the system is also used for acquiring a material vector related to the account vector in a material vector library based on the account vector, wherein the material vector is determined based on the incidence relation between materials and accounts;

and the recommending module is used for recommending the materials corresponding to the material vectors to the target account.

9. The apparatus of claim 8, wherein the apparatus further comprises: a determination module;

the determining module is used for acquiring reference data of an account, inputting the reference data into a skill tag prediction model and obtaining a first skill tag of the account;

10. The apparatus of claim 9, wherein the means for determining is configured to:

11. The apparatus of claim 9, wherein the means for determining is configured to:

12. The apparatus of claim 8, wherein the means for determining is configured to:

13. The apparatus of claim 12, wherein the means for determining is configured to:

14. The apparatus of claim 8, wherein the means for obtaining is configured to:

15. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

16. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-7.

17. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1-7.