CN115829653A - Method, device, equipment and medium for determining relevancy of advertisement text - Google Patents

Abstract

The disclosure provides a method, device, device and medium for determining the relevance of an advertisement text, and relates to the technical field of artificial intelligence, and in particular to the technical field of natural language processing. The implementation scheme is: acquiring first and second feature vector representations respectively corresponding to the first and second advertisement texts; and determining the correlation between the first and second advertisement texts based on the first and second feature vector representations degree, wherein the feature vector representation of the advertisement text is obtained by using the following determination process: obtaining historical advertisement data; obtaining revenue information of the advertisement text; based on the co-occurrence relationship between the multiple advertisement texts in the historical advertisement data, At least one associated text corresponding to each advertisement text; and based on the semantic and revenue information of each advertisement text and its corresponding associated text, determine the feature vector representation of each advertisement text.

Description

Advertisement Text Relevance Determination Method, Device, Equipment and Medium

technical field

The present disclosure relates to the field of artificial intelligence technology, in particular to the field of natural language processing, and in particular to a method, device, electronic device, computer-readable storage medium and computer program product for determining the relevance of advertisement text.

Background technique

Artificial intelligence is a discipline that studies the use of computers to simulate certain human thinking processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.), both at the hardware level and at the software level. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include computer vision technology, speech recognition technology, natural language processing technology, and machine learning/depth Learning, big data processing technology, knowledge map technology and other major directions.

In an advertising marketing scenario, it is usually necessary to actively recommend advertisements to users, especially to recommend advertisements related to the specific content to users based on specific content. Therefore, it is necessary to determine a recommendation strategy based on the relevance between the candidate content and the specific content.

The approaches described in this section are not necessarily approaches that have been previously conceived or employed. Unless otherwise indicated, it should not be assumed that any approaches described in this section are admitted to be prior art solely by virtue of their inclusion in this section. Similarly, issues mentioned in this section should not be considered to have been recognized in any prior art unless otherwise indicated.

Contents of the invention

The present disclosure provides a method, device, electronic device, computer-readable storage medium and computer program product for determining the relevance degree of advertisement text.

According to one aspect of the present disclosure, there is provided a method for determining the relevance of advertisement text, including: acquiring a first feature vector representation corresponding to the first advertisement text and a second feature vector representation corresponding to the second advertisement text; and based on The first feature vector representation and the second feature vector representation determine the correlation between the first advertisement text and the second advertisement text, wherein the first feature vector representation and the second feature vector representation are Obtained by the following determination process: obtaining historical advertisement data containing multiple advertisement texts, the multiple advertisement texts including the first advertisement text and the second advertisement text; obtaining each advertisement in the multiple advertisement texts Revenue information of the text; for each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, from at least one other advertisement text except the advertisement text Determine at least one associated text corresponding to the advertisement text in the text; and for each advertisement text in the plurality of advertisement texts, based on the semantics and income information of the advertisement text, and the corresponding semantics and The revenue information is to determine the feature vector representation corresponding to the advertisement text.

According to another aspect of the present disclosure, there is also provided a method for recommending an advertisement text, including: acquiring a target advertisement text and a plurality of candidate advertisement texts; the degree of relevance between the candidate advertisement texts; and based on the degree of relevance between the target advertisement text and the plurality of candidate advertisement texts, determine at least one advertisement text to be recommended from the plurality of candidate advertisement texts.

According to another aspect of the present disclosure, there is provided an apparatus for determining the relevance of advertisement text, including: a first acquisition unit configured to acquire the first feature vector representation corresponding to the first advertisement text and the first feature vector representation corresponding to the second advertisement text the second feature vector representation; and a first determining unit configured to determine the degree of relevance between the first advertisement text and the second advertisement text based on the first feature vector representation and the second feature vector representation , wherein, the first eigenvector representation and the second eigenvector representation are obtained by using a eigenvector determination unit, and the eigenvector determination unit includes: a first acquisition subunit configured to obtain a historical advertisement data, the multiple advertisement texts include the first advertisement text and the second advertisement text; the second acquisition subunit is configured to acquire the revenue information of each advertisement text in the multiple advertisement texts; the second A determining subunit, configured to, for each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, from at least Determining at least one associated text corresponding to the advertisement text from one other advertisement text; and the second determination subunit is configured to, for each advertisement text in the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text , and corresponding semantic and revenue information corresponding to at least one associated text, determine a feature vector representation corresponding to the advertisement text.

According to another aspect of the present disclosure, there is also provided an advertisement text recommendation device, including: a second acquisition unit configured to acquire a target advertisement text and a plurality of candidate advertisement texts; the above-mentioned apparatus for determining the relevance of advertisement texts, configured to determine the degree of relevance between the target advertisement text and the plurality of candidate advertisement texts; and a second determination unit configured to, based on the correlation between the target advertisement text and the plurality of candidate advertisement texts degree, at least one advertisement text to be recommended is determined from the plurality of candidate advertisement texts.

According to another aspect of the present disclosure, there is provided an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores information executable by the at least one processor. instructions, the instructions are executed by the at least one processor, so that the at least one processor can execute the above method for determining the relevance of advertisement text.

According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to make the computer execute the above method for determining the relevance of advertisement text.

According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein, when the computer program is executed by a processor, the above method for determining the relevance of an advertisement text can be realized.

According to one or more embodiments of the present disclosure, the degree of relevance between advertisement texts can be determined more accurately.

It should be understood that what is described in this section is not intended to identify key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will be readily understood through the following description.

Description of drawings

The drawings exemplarily illustrate the embodiment and constitute a part of the specification, and together with the text description of the specification, serve to explain the exemplary implementation of the embodiment. The illustrated embodiments are for illustrative purposes only and do not limit the scope of the claims. Throughout the drawings, like reference numbers designate similar, but not necessarily identical, elements.

FIG. 1 shows a schematic diagram of an exemplary system in which various methods described herein may be implemented according to an exemplary embodiment of the present disclosure;

FIG. 2 shows a flow chart of a process for determining a feature vector representation of an advertisement text according to an exemplary embodiment of the present disclosure;

Fig. 3 shows a flowchart of a method for determining the relevance of advertisement text according to an exemplary embodiment of the present disclosure;

FIG. 4 shows a schematic diagram of a text graph according to an exemplary embodiment of the present disclosure;

Fig. 5 shows a structural block diagram of a feature vector determining unit according to an exemplary embodiment of the present disclosure;

Fig. 6 shows a structural block diagram of an apparatus for determining the relevance of advertisement text according to an exemplary embodiment of the present disclosure;

FIG. 7 shows a structural block diagram of an exemplary electronic device that can be used to implement the embodiments of the present disclosure.

Detailed ways

Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and they should be regarded as exemplary only. 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 of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

In the present disclosure, unless otherwise stated, using the terms "first", "second", etc. to describe various elements is not intended to limit the positional relationship, temporal relationship or importance relationship of these elements, and such terms are only used for Distinguishes one element from another. In some examples, the first element and the second element may refer to the same instance of the element, and in some cases, they may also refer to different instances based on contextual description.

The terminology used in describing the various described examples in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly indicates otherwise, if the number of elements is not specifically limited, there may be one or more elements. In addition, the term "and/or" used in the present disclosure covers any one and all possible combinations of the listed items.

In related technologies, an implementation manner is to use the semantic similarity between advertisement texts to determine their relevance, or, use the co-occurrence relationship between advertisement texts in historical data to determine their relevance. However, these two methods have limited ability to express the correlation characteristics between multiple advertisement texts, and fail to take into account the revenue information corresponding to the advertisement texts.

Based on this, the present disclosure provides a method for determining the relevance of advertisement text, which uses the co-occurrence relationship between multiple advertisement texts in historical advertisement data to construct a text graph, and uses the information of each node and its associated associated nodes in the text graph to Semantic and revenue information, determine the feature vector representation of the advertising text corresponding to each node, and then use the feature vector representation to determine the correlation between the advertising texts, and can use the graph structure of the text graph to represent the text. On this basis, the revenue information is further combined to determine the correlation between advertisement texts to improve accuracy.

Embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

FIG. 1 shows a schematic diagram of an exemplary system 100 in which various methods and apparatus described herein may be implemented according to an embodiment of the present disclosure. 1, the system 100 includes one or more client devices 101, 102, 103, 104, 105, and 106, a server 120, and one or more communication networks coupling the one or more client devices to the server 120 110. Client devices 101, 102, 103, 104, 105, and 106 may be configured to execute one or more applications.

In an embodiment of the present disclosure, the server 120 may run one or more services or software applications that enable the execution of the method for determining the relevance of advertisement text.

In some embodiments, server 120 may also provide other services or software applications, which may include non-virtual environments and virtual environments. In some embodiments, these services may be provided as web-based services or cloud services, such as under a software-as-a-service (SaaS) model to users of client devices 101, 102, 103, 104, 105, and/or 106 .

In the configuration shown in FIG. 1 , server 120 may include one or more components that implement the functions performed by server 120 . These components may include software components, hardware components or combinations thereof executable by one or more processors. Users operating client devices 101, 102, 103, 104, 105, and/or 106 may in turn utilize one or more client application programs to interact with server 120 to utilize the services provided by these components. It should be understood that various different system configurations are possible, which may differ from system 100 . Accordingly, FIG. 1 is one example of a system for implementing the various methods described herein, and is not intended to be limiting.

A user may use client devices 101, 102, 103, 104, 105, and/or 106 to send advertising text. A client device may provide an interface that enables a user of the client device to interact with the client device. The client device can also output information to the user via the interface. Although FIG. 1 depicts only six client devices, those skilled in the art will understand that the present disclosure can support any number of client devices.

Client devices 101, 102, 103, 104, 105, and/or 106 may include various classes of computing devices, such as portable handheld devices, general-purpose computers (such as personal computers and laptops), workstation computers, wearable devices, Smart screen devices, self-service terminal devices, service robots, game systems, thin clients, various messaging devices, sensors or other sensing devices, etc. These computer devices can run various classes and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, Linux or Linux-like operating systems (such as GOOGLE Chrome OS); or include various mobile operating systems , such as MICROSOFT WindowsMobile OS, iOS, Windows Phone, Android. Portable handheld devices may include cellular phones, smart phones, tablet computers, personal digital assistants (PDAs), and the like. Wearable devices may include head-mounted displays (such as smart glasses) and other devices. Gaming systems may include various handheld gaming devices, Internet-enabled gaming devices, and the like. A client device is capable of executing a variety of different applications, such as various Internet-related applications, communication applications (eg, email applications), Short Message Service (SMS) applications, and may use a variety of communication protocols.

Network 110 can be any type of network known to those skilled in the art that can support data communications using any of a number of available protocols, including but not limited to TCP/IP, SNA, IPX, and the like. By way of example only, the one or more networks 110 may be a local area network (LAN), an Ethernet-based network, a token ring, a wide area network (WAN), the Internet, a virtual network, a virtual private network (VPN), an intranet, an extranet, Blockchain network, public switched telephone network (PSTN), infrared network, wireless network (eg Bluetooth, WIFI) and/or any combination of these and/or other networks.

Server 120 may include one or more general purpose computers, dedicated server computers (e.g., PC (personal computer) servers, UNIX servers, midrange servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and/or combination . Server 120 may include one or more virtual machines running virtual operating systems, or other computing architectures involving virtualization (eg, one or more flexible pools of logical storage devices that may be virtualized to maintain the server's virtual storage devices). In various embodiments, server 120 may run one or more services or software applications that provide the functionality described below.

Computing units in server 120 may run one or more operating systems including any of the operating systems described above as well as any commercially available server operating systems. Server 120 may also run any of a variety of additional server applications and/or middle-tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, and the like.

In some implementations, server 120 may include one or more applications to analyze and consolidate data feeds and/or event updates received from users of client devices 101 , 102 , 103 , 104 , 105 , and 106 . Server 120 may also include one or more applications to display data feeds and/or real-time events via one or more display devices of client devices 101 , 102 , 103 , 104 , 105 , and 106 .

In some implementations, the server 120 may be a server of a distributed system, or a server combined with blockchain. The server 120 can also be a cloud server, or an intelligent cloud computing server or an intelligent cloud host with artificial intelligence technology. Cloud server is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability existing in traditional physical host and virtual private server (VPS, Virtual Private Server) services.

System 100 may also include one or more databases 130 . In some embodiments, these databases may be used to store data and other information. For example, one or more of databases 130 may be used to store information such as audio files and video files. Database 130 may reside in various locations. For example, the database used by server 120 may be local to server 120, or may be remote from server 120 and may communicate with server 120 via a network-based or dedicated connection. Database 130 can be of different kinds. In some embodiments, the database used by server 120 may be, for example, a relational database. One or more of these databases may store, update and retrieve the database and data from the database in response to commands.

In some embodiments, one or more of databases 130 may also be used by applications to store application data. Databases used by applications can be of different classes such as key-value stores, object stores or regular stores backed by a file system.

The system 100 of FIG. 1 may be configured and operated in various ways to enable application of the various methods and apparatuses described in accordance with this disclosure.

Fig. 2 shows a flowchart of a process 200 for determining a feature vector representation of an advertisement text according to an exemplary embodiment of the present disclosure. As shown in Figure 2, the process 200 includes:

Step S201, acquiring historical advertisement data including a plurality of advertisement texts, the plurality of advertisement texts including the first advertisement text and the second advertisement text;

Step S202, obtaining the revenue information of each advertisement text in the plurality of advertisement texts;

Step S203, for each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, from at least one other advertisement text except the advertisement text determining at least one associated text corresponding to the advertisement text; and

Step S204, for each advertisement text in the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text and the semantics and revenue information corresponding to at least one associated text, determine the feature vector corresponding to the advertisement text express.

Fig. 3 shows a flowchart of a method 300 for determining the relevance of advertisement text according to an exemplary embodiment of the present disclosure. As shown in FIG. 3, method 300 includes:

Step S301. Obtain a first feature vector representation corresponding to the first advertisement text and a second feature vector representation corresponding to the second advertisement text, wherein the first feature vector representation and the second feature vector representation are obtained by using the process 200 of; and

Step S302, based on the first feature vector representation and the second feature vector representation, determine the degree of correlation between the first advertisement text and the second advertisement text.

Therefore, based on the above-mentioned determination process 200, the co-occurrence relationship between multiple advertisement texts in the historical advertisement data can be used to determine the associated text corresponding to each advertisement text, and then based on the semantics and benefits of each advertisement text and its associated text Information, to determine the feature vector representation of each advertisement text, so that the feature vector representation of the advertisement text can more accurately characterize the correlation between the advertisement text and other advertisement texts, and at the same time represent the revenue information of the advertisement text. Furthermore, by using the above-mentioned feature vector representation to determine the correlation between advertisement texts, it is possible to determine the correlation between advertisement texts based on the correlation between advertisement texts indicated by the text co-occurrence relationship and further combining revenue information , to improve accuracy.

In some examples, the advertisement text may be, for example, the text of the advertisement name recommended to the user, or the text of the product name actively queried by the user, as long as it can represent the relevant information of the advertisement marketing object, which is not limited in the present disclosure.

In some examples, when the advertisement text is the advertisement title text recommended to the user, the revenue information of the advertisement text may be determined based on the advertisement cost. It should be understood that the revenue information of the advertisement text may also include other content, for example, it may be the revenue information corresponding to the marketing object represented by the advertisement text, such as product sales, page views, video playback volume, etc., which is not limited in this disclosure .

In some examples, determining the correlation between the first advertisement text and the second advertisement text based on the first feature vector representation and the second feature vector representation in step S302 may include: based on the first feature vector represents the similarity between the representation and the second feature vector representation, and determines the correlation between the first advertisement text and the second advertisement text. The similarity may be determined, for example, by calculating the Euclidean distance between the first feature vector representation and the second feature vector representation, or may also be determined by calculating cosine similarity, Manhattan distance and other indicators.

In some examples, after using the process 200 to determine the feature vector representations of multiple advertisement texts, the multiple advertisement texts and the corresponding multiple feature vector representations can be stored in the database, so that in step S301 based on the first advertisement text and For the second advertisement text, the corresponding first feature vector representation and the second feature vector representation are queried from the database, so as to improve the efficiency of determining the relevance of the advertisement text.

According to some embodiments, for each advertisement text in the plurality of advertisement texts in step S203, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, at least Determining at least one associated text corresponding to the advertising text in one other advertising text includes: based on the co-occurrence relationship between the multiple advertising texts in the historical advertising data, constructing a one-to-one correspondence with the multiple advertising texts A text graph of a plurality of nodes, wherein, for any two advertisement texts of the plurality of texts, in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, based on the two The node corresponding to the advertisement text establishes a connection edge; and for each node in the plurality of nodes, based on the connection relationship between the plurality of nodes, at least one associated node corresponding to the node is determined from at least one other node , to obtain at least one associated text corresponding to the corresponding advertisement text.

Thus, based on the above determination process 200, the co-occurrence relationship between multiple advertisement texts in the historical advertisement data can be used to construct a text graph, and the semantic and revenue information of each node and its associated associated nodes in the text graph can be used to determine each The feature vector representation of the advertising text corresponding to each node, so that the feature vector representation of the advertising text can more accurately characterize the correlation between the advertising text and other advertising texts, and at the same time represent the revenue information of the advertising text. Further, by using the above-mentioned feature vector representation to determine the correlation between advertisement texts, it is possible to determine the correlation between advertisement texts based on the association relationship between texts represented by the graph structure of the text graph and further combining revenue information , to improve accuracy.

According to some embodiments, for each advertisement text in the plurality of advertisement texts in step 204, based on the semantics and revenue information of the advertisement text and the corresponding semantics and revenue information of at least one associated text, determine the The feature vector representation includes: for each connecting edge in at least one connecting edge between the node corresponding to the advertising text and the corresponding at least one associated node, based on the semantics of the advertising text respectively corresponding to the two endpoints of the connecting edge and revenue information, determine an edge vector representation of the connected edge; and determine a feature vector representation of the advertisement text based on the edge vector representation of the at least one connected edge. Therefore, the edge vector representation can be used to characterize the relevance features of the advertisement text corresponding to the two endpoints of each edge, and then the multiple edge vectors between each node and its associated nodes can be aggregated to obtain a more accurate The feature vector representation of the text corresponding to each node.

In some examples, the determining the feature vector representation of the advertisement text corresponding to the node based on the edge vector representation of the at least one connection edge may include: calculating an average value of the edge vector representations of the at least one connection edge to obtain A feature vector representation of the advertisement text corresponding to the node is obtained. This step can also be performed in other ways, for example, it can be calculated based on the edge vector representation of the at least one connection edge and a preset formula, so as to obtain the feature vector representation of the advertisement text corresponding to the node.

According to some embodiments, the determining the edge vector representation of the connecting edge based on the semantics and revenue information of the advertising text corresponding to the two endpoints of the connecting edge includes: determining the relationship between the advertising text corresponding to the two endpoints respectively ; and based on the semantic similarity and the revenue information of the advertising text corresponding to the two endpoints, determine the edge vector representation of the connecting edge. Therefore, the semantic similarity can be used to represent the correlation between the advertisement texts corresponding to the two endpoints, which is more convenient and accurate.

In some examples, it may be to determine the semantic feature vector representations of the advertisement text corresponding to the two endpoints, and then determine the semantic similarity based on the similarity between the two semantic feature vector representations. The semantic feature vector representation may be realized, for example, by inputting advertisement text into a language model, or may also be realized by querying a database storing multiple texts and corresponding semantic feature vectors, which is not limited in the present disclosure.

According to some embodiments, the determining the semantic similarity between the advertisement texts respectively corresponding to the two endpoints includes: inputting the advertisement texts respectively corresponding to the two endpoints of the connecting edge into a pre-trained language model to obtain the The semantic similarity output by a pre-trained language model, wherein the pre-trained language model is obtained by using labeled corpus data for training. As a result, the semantic similarity between two advertisement texts can be obtained by using the pre-trained language model, improving efficiency and accuracy.

In some examples, the pre-trained language model may be an Ernie model, for example.

According to some embodiments, the determining process 200 further includes: for each connecting edge in at least one connecting edge between the node corresponding to the advertisement text and the corresponding at least one associated node, determining two of the connecting edges The co-occurrence frequencies between the advertising texts corresponding to the endpoints, and wherein, based on the semantics and revenue information of the advertising texts corresponding to the two endpoints of the connecting edge and the co-occurrence frequency, the edge vector representation of the connecting edge is determined . Thus, the co-occurrence frequency of the advertisement text corresponding to the two endpoints can be combined to more accurately indicate the correlation between the two endpoints.

In some examples, the co-occurrence frequency may be the number of co-occurrence cnt. Therefore, the fusion information S of the number of co-occurrence cnt and the income information acp can be determined based on the following formula:

S=a*cnt+b*acp

In this example, a and b in the formula may be weight values set according to actual needs.

According to some embodiments, the determining process 200 further includes: respectively performing normalization processing on the co-occurrence frequency and the revenue information to obtain the normalized co-occurrence frequency and the normalized revenue information, and wherein, based on The two endpoints of the connecting edge correspond to the semantics of the advertisement text, the normalized revenue information and the normalized co-occurrence frequency respectively, and determine the edge vector representation of the connecting edge. As a result, the numerical ranges of co-occurrence frequency and revenue information can be scaled to the same interval, the calculation process can be simplified, and a more accurate edge vector representation can be obtained.

In some examples, normalization may be performed using the following formula:

In the formula, x is the initial revenue information or co-occurrence frequency, x _min is the minimum value of revenue information or co-occurrence frequency in all data, x _max is the maximum value of revenue information or co-occurrence frequency in all data, and X′ is the regression Normalized return information or normalized co-occurrence frequency.

According to some embodiments, determining the edge vector representation of the connected edge based on the semantics and revenue information of the advertisement text corresponding to the two endpoints of the connected edge includes: assigning the advertisement corresponding to the two endpoints of the connected edge The text and its revenue information are input into the edge vector encoding model to obtain the edge vector representation output by the edge vector encoding model, wherein the edge vector encoding model is obtained by training in the following manner: obtaining information including multiple A sample text graph of a plurality of nodes corresponding to the sample text one-to-one and revenue information of each sample text in the plurality of sample texts, the sample text graph includes a plurality of connections for connecting the plurality of nodes side; for each of the multiple connection edges included in the sample text graph, input the sample text corresponding to the two end points of the connection edge and its income information into the edge vector coding model to obtain the The edge vector representation of the connecting plate output by the edge vector encoding model; based on the edge vector representations of the plurality of connection edges, determine the feature vector representations of a plurality of sample texts corresponding to the plurality of nodes; obtain the plurality of The real correlation between the first sample text and the second sample text in the sample text; based on the first feature vector representation and the second feature vector representation respectively corresponding to the first sample text and the second sample text, determining a predicted correlation between the first sample text and the second sample text; and adjusting parameters of the edge vector coding model based on the real correlation and the predicted correlation. In this way, the side vector representation can be obtained by using the side vector coding model, and the correlation prediction task between texts can be performed using the side vector representation output by the side vector coding model, and model training can be carried out based on this to improve efficiency and accuracy.

In some examples, the co-occurrence frequencies of the advertising texts corresponding to the two endpoints of the connecting edge can also be input into the edge vector coding model at the same time, for example, the fusion information S described above can be input into the edge vector coding model , to obtain an edge vector representation that can more accurately characterize the correlation between two advertisement texts.

According to some embodiments, the plurality of advertisement texts include a plurality of historical query texts and a plurality of historical recommendation texts, the historical advertisement data includes a plurality of text pairs, each text pair in the plurality of text pairs includes a historical query texts and historical recommended texts recommended to users based on the historical query texts, and in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, the corresponding The node establishing an edge of the text graph for the vertex includes: in response to determining that the historical advertisement data includes a text pair composed of the two advertisement texts, establishing an edge of the text graph with nodes corresponding to the two advertisement texts as vertices. Therefore, the correlation represented by the mapping relationship between the historical query text and the historical recommended text in the historical advertising data can be fully utilized, and the edges of the text graph can be constructed based on this, so that the edges in the text graph can accurately indicate the relationship between the two endpoint nodes. relevance.

In some examples, the co-occurrence frequency between corresponding two advertisement texts may be determined based on the frequency of each text pair in the historical advertisement data.

According to some embodiments, for each node in the text graph, determining at least one associated node corresponding to the node from multiple other nodes based on the connection relationship between the multiple nodes includes: for all For each of the plurality of other nodes, in response to the number of connection hops between the other node and the node being not greater than a preset threshold, the other node is determined to be the associated node. In this way, other nodes closer to the node can be determined as associated nodes, preventing distant nodes from affecting the feature vector representation of the node's advertisement text, and improving accuracy.

According to some embodiments, for each node in the text graph, determining at least one associated node corresponding to the node from multiple other nodes based on the connection relationship between the multiple nodes includes: based on the According to the connection relationship between the plurality of nodes and preset rules, determine the respective sampling probabilities of the plurality of other nodes, wherein, according to the preset rules, the connection between the plurality of other nodes and the node The sampling probability of a node with a smaller hop count is greater than the sampling probability of a node with a larger connection hop count between the node; and based on the sampling probability, random sampling is performed on the plurality of other nodes to obtain the At least one associated node. Therefore, the associated nodes corresponding to each node can be obtained by using hierarchical random sampling, and the closer the node is, the higher the sampling probability is, which reduces the number of associated nodes and simplifies the vector calculation process while ensuring accuracy.

Fig. 4 shows a schematic diagram of a text graph according to an exemplary embodiment of the present disclosure. In this example, for node a at the center, for example, two layers of neighbor nodes associated with it can be sampled, specifically, nodes b, c, and d whose connection hops to a are one are the first layer nodes, Nodes e, f, and g whose connection hops with a are two are the second-layer nodes. The sampling frequency corresponding to each layer of nodes may decrease as the number of layers increases, for example, the sampling frequency corresponding to each layer of nodes may be determined based on an exponential decay function.

In this example, by utilizing the steps described above, edge vector representations of edges a-b, edge a-c, edge a-d, edge b-e, edge d-f, and edge d-g can be obtained respectively. Therefore, the nodes of each layer can be aggregated separately, that is, the layer is determined based on the edge vector representation of the connection edge between the node of this layer and the node of the next layer and the corresponding feature vector representation of the node of the next layer The eigenvector representation corresponding to the nodes of . For example, the eigenvector representation corresponding to node d can be determined based on the edge vector representation of edge d-f and edge d-g, and then the node a can be determined based on the edge vector representation of edge a-b, edge a-c, edge a-d and the eigenvector representation of nodes b, c, and d A feature vector representation of the corresponding advertisement text.

According to another aspect of the present disclosure, there is also provided a method for recommending an advertisement text, including: acquiring a target advertisement text and a plurality of candidate advertisement texts; the degree of relevance between the candidate advertisement texts; and based on the degree of relevance between the target advertisement text and the plurality of candidate advertisement texts, determine at least one advertisement text to be recommended from the plurality of candidate advertisement texts.

In some examples, the multiple candidate advertisement texts may be sorted based on the correlation between the target advertisement text and the multiple candidate advertisement texts, and at least one advertisement text to be recommended is determined based on the sorting result.

According to another aspect of the present disclosure, an apparatus for determining the relevance of advertisement text is also provided. Fig. 5 shows a structural block diagram of a feature vector determining unit 500 according to an exemplary embodiment of the present disclosure. Fig. 6 shows a structural block diagram of an apparatus 600 for determining the relevance of advertisement text according to an exemplary embodiment of the present disclosure.

As shown in Figure 5, the feature vector determination unit 500 includes:

The first acquiring subunit 501 is configured to acquire historical advertisement data including a plurality of advertisement texts, the plurality of advertisement texts including the first advertisement text and the second advertisement text;

The second obtaining subunit 502 is configured to obtain the revenue information of each advertisement text in the plurality of advertisement texts;

The first determination subunit 503 is configured to, for each advertisement text in the multiple advertisement texts, based on the co-occurrence relationship among the multiple advertisement texts in the historical advertisement data, select at least one associated text that corresponds to the ad text identified in at least one other ad text; and

The second determination subunit 504 is configured to, for each advertisement text in the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text and the semantics and revenue information corresponding to at least one associated text, determine the corresponding The feature vector representation corresponding to the advertisement text.

The operations of unit 501 - unit 504 of unit 500 are similar to the operations of step S201 - step S204 described above, and will not be repeated here.

As shown in FIG. 6 , the apparatus 600 for determining the relevance of advertisement text includes:

The first acquiring unit 601 is configured as a first feature vector representation corresponding to the first advertisement text and a second feature vector representation corresponding to the second advertisement text; and

The first determining unit 602 is configured to determine the correlation between the first advertisement text and the second advertisement text based on the first feature vector representation and the second feature vector representation, wherein the first feature vector The representation and the second eigenvector representation are obtained using unit 200 .

The operations of unit 601 - unit 602 of the device 600 are similar to the operations of step S301 - step S302 described above, and will not be repeated here.

According to some embodiments, the first determination subunit includes: a construction module configured to, based on the co-occurrence relationship among the multiple advertisement texts in the historical advertisement data, construct an A text graph corresponding to a plurality of nodes, wherein, for any two advertisement texts of the plurality of texts, in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, based on the The nodes corresponding to the two advertisement texts establish a connection edge; and the first determination module is configured to, for each node in the plurality of nodes, based on the connection relationship between the plurality of nodes, from at least one other node Determine at least one associated node corresponding to the node, so as to obtain at least one associated text corresponding to the corresponding advertisement text.

According to some embodiments, the second determination subunit includes: a second determination module configured to, for each connection edge in at least one connection edge between the node corresponding to the advertisement text and the corresponding at least one associated node , based on the semantics and revenue information of the advertisement text corresponding to the two endpoints of the connecting edge, determine the edge vector representation of the connecting edge; and a third determining module, configured to be based on the edge vector representation of the at least one connecting edge , to determine the feature vector representation of the advertisement text.

According to some embodiments, the second determination module is configured to: determine the semantic similarity between the advertisement texts respectively corresponding to the two endpoints; and determine the advertisement text corresponding to the two endpoints based on the semantic similarity The revenue information of the text determines the edge vector representation of the connected edge.

According to some embodiments, the second determination module is configured to: input the advertisement texts respectively corresponding to the two endpoints of the connecting edge into a pre-trained language model, so as to obtain the semantic similarity outputted by the pre-trained language model degree, wherein the pre-trained language model is obtained by using labeled corpus data for training.

According to some embodiments, the feature vector determining unit further includes: a third determining subunit configured to, for each connection in at least one connection edge between the node corresponding to the advertisement text and the corresponding at least one associated node edge, determining the co-occurrence frequency between the advertisement texts respectively corresponding to the two endpoints of the connecting edge, and wherein, the second determination module is configured to be based on the Semantic and revenue information, as well as the co-occurrence frequency, determine the edge vector representation of the connected edge.

According to some embodiments, the eigenvector determining unit further includes: a processing subunit configured to perform normalization processing on the co-occurrence frequency and the income information respectively to obtain a normalized co-occurrence frequency and a normalized revenue information, and wherein the first determining module is configured to determine the Edge vector representation of connected edges.

According to some embodiments, the second determination module is configured to: input the advertisement text corresponding to the two endpoints of the connecting edge and its revenue information into the edge vector encoding model, so as to obtain the output of the edge vector encoding model The edge vector representation, wherein the edge vector coding model is obtained by training in the following manner: obtain a sample text graph containing a plurality of nodes corresponding to a plurality of sample texts one-to-one and each of the plurality of sample texts Revenue information of a sample text graph, the sample text graph includes a plurality of connection edges used to connect the plurality of nodes; for each connection edge in the plurality of connection edges included in the sample text graph, the The sample text corresponding to the two endpoints of the connection edge and the income information thereof are input into the edge vector encoding model to obtain the edge vector representation of the connection plate output by the edge vector encoding model; based on the multiple connections The edge vector representation of the edge determines the feature vector representation of a plurality of sample texts corresponding to the plurality of nodes; obtains the true correlation between the first sample text and the second sample text in the plurality of sample texts; based on The first feature vector representation and the second feature vector representation corresponding to the first sample text and the second sample text respectively, determine the predicted correlation between the first sample text and the second sample text; and based on The real correlation degree and the predicted correlation degree adjust parameters of the side vector coding model.

According to some embodiments, the plurality of advertisement texts include a plurality of historical query texts and a plurality of historical recommendation texts, the historical advertisement data includes a plurality of text pairs, each text pair in the plurality of text pairs includes a historical query text and historical recommended text recommended to users based on the historical query text, and wherein the building block is configured to: in response to determining that the historical advertisement data includes a text pair consisting of the two advertisement texts, The edges of the text graph are established with the nodes corresponding to the two advertisement texts as vertices.

According to some embodiments, the first determination module is configured to: for each other node among the plurality of other nodes, in response to the number of connection hops between the other node and the node being not greater than a preset threshold, Determine the other node as the associated node.

According to some embodiments, the first determination module is configured to: determine the respective sampling probabilities of the plurality of other nodes based on the connection relationship between the plurality of nodes and preset rules, wherein, according to the preset rule, the sampling probability of a node with a smaller connection hop between the nodes among the plurality of other nodes is greater than the sampling probability of a node with a larger connection hop between the node; and based on the sampling probability, performing random sampling on the plurality of other nodes to obtain the at least one associated node.

According to another aspect of the present disclosure, there is also provided an advertisement text recommendation device, including: a second acquisition unit configured to acquire a target advertisement text and a plurality of candidate advertisement texts; the above-mentioned apparatus for determining the relevance of advertisement texts, configured to determine the degree of relevance between the target advertisement text and the plurality of candidate advertisement texts; and a second determination unit configured to, based on the correlation between the target advertisement text and the plurality of candidate advertisement texts degree, at least one advertisement text to be recommended is determined from the plurality of candidate advertisement texts.

According to another aspect of the present disclosure, there is also provided an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores information executable by the at least one processor. instructions, the instructions are executed by the at least one processor, so that the at least one processor can execute the above-mentioned method for determining the relevance of advertisement text.

According to another aspect of the present disclosure, there is also provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to make the computer execute the above-mentioned method for determining the relevance of advertisement text.

According to another aspect of the present disclosure, there is also provided a computer program product, including a computer program, wherein, when the computer program is executed by a processor, the above-mentioned method for determining the relevance of an advertisement text is implemented.

Referring to FIG. 7 , a structural block diagram of an electronic device 700 that can serve as a server or a client of the present disclosure will now be described, which is an example of a hardware device that can be applied to various aspects of the present disclosure. Electronic device is intended to mean various forms of digital electronic computing equipment, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are by way of example only, and are not intended to limit implementations of the disclosure described and/or claimed herein.

As shown in FIG. 7, the device 700 includes a computing unit 701 that can execute according to a computer program stored in a read-only memory (ROM) 702 or loaded from a storage unit 708 into a random-access memory (RAM) 703. Various appropriate actions and treatments. In the RAM 703, various programs and data necessary for the operation of the device 700 can also be stored. The computing unit 701 , ROM 702 , and RAM 703 are connected to each other through a bus 704 . An input/output (I/O) interface 705 is also connected to the bus 704 .

Multiple components in the device 700 are connected to the I/O interface 705 , including: an input unit 706 , an output unit 707 , a storage unit 708 and a communication unit 709 . The input unit 706 may be any type of device capable of inputting information to the device 700, the input unit 706 may receive input digital or character information, and generate key signal input related to user settings and/or function control of the electronic device, and may Including but not limited to mouse, keyboard, touch screen, trackpad, trackball, joystick, microphone and/or remote control. The output unit 707 may be any kind of device capable of presenting information, and may include, but is not limited to, a display, a speaker, a video/audio output terminal, a vibrator, and/or a printer. The storage unit 708 may include, but is not limited to, a magnetic disk and an optical disk. The communication unit 709 allows the device 700 to exchange information/data with other devices via a computer network such as the Internet and/or various telecommunication networks, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver and/or a chipset , such as Bluetooth™ devices, 802.11 devices, WiFi devices, WiMax devices, cellular communication devices, and/or the like.

The computing unit 701 may be various general-purpose and/or special-purpose processing components having processing and computing capabilities. Some examples of computing units 701 include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processing processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 executes various methods and processes described above, such as a method for determining the relevance of advertisement text. For example, in some embodiments, the method for determining the relevance of advertisement text may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 708 . In some embodiments, part or all of the computer program may be loaded and/or installed on the device 700 via the ROM 702 and/or the communication unit 709 . When the computer program is loaded into RAM 703 and executed by computing unit 701, one or more steps of the method for determining the relevance of advertisement text described above can be executed. Alternatively, in other embodiments, the computing unit 701 may be configured in any other appropriate way (for example, by means of firmware) to execute the method for determining the relevance of advertisement text.

Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips Implemented in a system of systems (SOC), complex programmable logic device (CPLD), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include being implemented in one or more computer programs executable and/or interpreted on a programmable system including at least one programmable processor, the programmable processor Can be special-purpose or general-purpose programmable processor, can receive data and instruction from storage system, at least one input device, and at least one output device, and transmit data and instruction to this storage system, this at least one input device, and this at least one output device an output device.

Program codes 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, a special purpose computer, or other programmable data processing devices, so that the program codes, when executed by the processor or controller, make the functions/functions specified in the flow diagrams and/or block diagrams Action is implemented. 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 the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A 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, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media would include one or more wire-based electrical connections, portable computer discs, hard drives, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, compact disk read only memory (CD-ROM), optical storage, magnetic storage, or any suitable combination of the foregoing.

To provide for interaction with the user, the systems and techniques described herein 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 the user. ); and a keyboard and pointing device (eg, a mouse or a trackball) through which a user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and can be in any form (including Acoustic input, speech input or, tactile input) to receive input from the user.

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

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

It should be understood that steps may be reordered, added or deleted using the various forms of flow shown above. For example, each step described in the present disclosure may be executed in parallel, sequentially or in a different order, as long as the desired result of the technical solution disclosed in the present disclosure can be achieved, no limitation is imposed herein.

Although the embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it should be understood that the above-mentioned methods, systems and devices are merely exemplary embodiments or examples, and the scope of the present invention is not limited by these embodiments or examples, but It is limited only by the appended claims and their equivalents. Various elements in the embodiments or examples may be omitted or replaced by equivalent elements thereof. Also, steps may be performed in an order different from that described in the present disclosure. Further, various elements in the embodiments or examples can be combined in various ways. Importantly, as technology advances, many of the elements described herein may be replaced by equivalent elements appearing after this disclosure.

Claims (27)

1. A method for determining the relevance of an advertisement text, comprising: acquiring a first feature vector representation corresponding to the first advertisement text and a second feature vector representation corresponding to the second advertisement text; and determining a degree of relevance between the first advertisement text and the second advertisement text based on the first feature vector representation and the second feature vector representation, Wherein, the first eigenvector representation and the second eigenvector representation are obtained using the following determination process: acquiring historical advertisement data including a plurality of advertisement texts, the plurality of advertisement texts including the first advertisement text and the second advertisement text; Acquiring revenue information of each advertisement text in the plurality of advertisement texts; For each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, determine from at least one other advertisement text other than the advertisement text that is related to the advertisement text at least one associated text to which the advertisement text corresponds; and For each advertisement text in the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text and the semantics and revenue information corresponding to at least one associated text, determine the feature vector representation corresponding to the advertisement text. 2. The method according to claim 1, wherein, for each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, from The at least one associated text determined to correspond to the advertisement text in at least one other advertisement text other than the advertisement text includes: Based on the co-occurrence relationship among the plurality of advertisement texts in the historical advertisement data, construct a text graph containing a plurality of nodes corresponding to the plurality of advertisement texts one-to-one, wherein, for the plurality of texts Any two advertisement texts, in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, establish a connection edge based on the nodes corresponding to the two advertisement texts; and For each node in the plurality of nodes, based on the connection relationship between the plurality of nodes, determine at least one associated node corresponding to the node from at least one other node, so as to obtain the corresponding advertisement text At least one associated text. 3. The method according to claim 2, wherein, for each advertisement text in the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text, and the corresponding semantics and revenue information of at least one associated text Information to determine the feature vector representation of the advertisement text includes: For each connection edge in at least one connection edge between the node corresponding to the advertisement text and the corresponding at least one associated node, based on the semantics and revenue information of the advertisement text corresponding to the two endpoints of the connection edge, determine an edge vector representation of the connected edge; and Based on the edge vector representation of the at least one connected edge, a feature vector representation of the advertisement text is determined. 4. The method according to claim 3, wherein said determining the edge vector representation of the connecting edge based on the semantics and revenue information of the advertising text respectively corresponding to the two endpoints of the connecting edge comprises: determining the semantic similarity between the advertisement texts respectively corresponding to the two endpoints; and Based on the semantic similarity and the revenue information of the advertisement text corresponding to the two endpoints, determine the edge vector representation of the connecting edge. 5. The method according to claim 4, wherein said determining the semantic similarity between the advertisement texts corresponding to the two endpoints comprises: Inputting the advertisement texts corresponding to the two endpoints of the connecting edge into the pre-training language model to obtain the semantic similarity output by the pre-training language model, wherein the pre-training language model utilizes labeled corpus data obtained by training. 6. The method according to any one of claims 3-5, wherein the determining process further comprises: For each connection edge in at least one connection edge between the node corresponding to the advertisement text and the corresponding at least one associated node, determine the co-occurrence frequency between the advertisement texts respectively corresponding to the two endpoints of the connection edge, And wherein, based on the semantics and revenue information of the advertisement text corresponding to the two end points of the connecting edge and the co-occurrence frequency, the edge vector representation of the connecting edge is determined. 7. The method of claim 6, wherein the determining process further comprises: Performing normalization processing on the co-occurrence frequency and the revenue information respectively to obtain a normalized co-occurrence frequency and a normalized revenue information, And wherein, based on the semantics and normalized revenue information of the advertisement text corresponding to the two endpoints of the connecting edge and the normalized co-occurrence frequency, the edge vector representation of the connecting edge is determined. 8. The method according to any one of claims 2-7, wherein, based on the semantics and revenue information of the advertising text corresponding to the two endpoints of the connecting edge, determining the edge vector representation of the connecting edge includes : inputting the advertisement text corresponding to the two endpoints of the connecting edge and its revenue information into the edge vector encoding model to obtain the edge vector representation output by the edge vector encoding model, Wherein, the edge vector coding model is obtained by training in the following manner: Obtaining a sample text graph containing a plurality of nodes one-to-one corresponding to a plurality of sample texts and revenue information of each sample text in the plurality of sample texts, the sample text graph includes multiple Multiple connection edges of nodes; For each connection edge among the plurality of connection edges included in the sample text graph, input the sample text corresponding to the two end points of the connection edge and its revenue information into the edge vector encoding model to obtain the edge vector an edge vector representation of said connected plate output by the encoding model; Based on the edge vector representations of the plurality of connection edges, determine feature vector representations of a plurality of sample texts corresponding to the plurality of nodes; Acquiring the real correlation between the first sample text and the second sample text among the plurality of sample texts; determining a predicted correlation between the first sample text and the second sample text based on the first feature vector representation and the second feature vector representation respectively corresponding to the first sample text and the second sample text; and Adjusting parameters of the edge vector coding model based on the true correlation and the predicted correlation. 9. The method according to any one of claims 2-8, wherein said plurality of advertisement texts comprises a plurality of historical query texts and a plurality of historical recommendation texts, said historical advertisement data comprises a plurality of text pairs, so Each of the plurality of text pairs includes a historical query text and historical recommendation text recommended to users based on the historical query text, And wherein, in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, establishing an edge of the text graph with nodes corresponding to the two advertisement texts as vertices includes: In response to determining that the historical advertisement data includes a text pair consisting of the two advertisement texts, an edge of the text graph is established with nodes corresponding to the two advertisement texts as vertices. 10. The method according to any one of claims 2-9, wherein, for each node in the text graph, based on the connection relationship between the plurality of nodes, from a plurality of other nodes Determining at least one associated node corresponding to the node includes: For each of the plurality of other nodes, in response to the number of connection hops between the other node and the node being not greater than a preset threshold, it is determined that the other node is the associated node. 11. The method according to any one of claims 2-9, wherein, for each node in the text graph, based on the connection relationship between the plurality of nodes, from a plurality of other nodes Determining at least one associated node corresponding to the node includes: Based on the connection relationship between the plurality of nodes and preset rules, determine the respective sampling probabilities of the plurality of other nodes, wherein, according to the preset rules, among the plurality of other nodes and between the nodes The sampling probability of a node with a smaller connection hop is greater than the sampling probability of a node with a larger connection hop between the node; and Randomly sampling the plurality of other nodes based on the sampling probability to obtain the at least one associated node. 12. A method for recommending an advertisement text, comprising: Obtain the target advertisement text and multiple candidate advertisement texts; Using the method according to any one of claims 1-11, determining the degree of relevance between the target advertisement text and the plurality of candidate advertisement texts; and Based on the correlation between the target advertisement text and the plurality of candidate advertisement texts, at least one advertisement text to be recommended is determined from the plurality of candidate advertisement texts. 13. A device for determining the relevance of an advertisement text, comprising: A first acquiring unit configured to acquire a first feature vector representation corresponding to the first advertisement text and a second feature vector representation corresponding to the second advertisement text; and The first determining unit is configured to determine the correlation between the first advertisement text and the second advertisement text based on the first feature vector representation and the second feature vector representation, Wherein, the first eigenvector representation and the second eigenvector representation are obtained by using a eigenvector determination unit, and the eigenvector determination unit includes: The first acquisition subunit is configured to acquire historical advertisement data including a plurality of advertisement texts, the plurality of advertisement texts including the first advertisement text and the second advertisement text; The second obtaining subunit is configured to obtain the revenue information of each advertisement text in the plurality of advertisement texts; The first determining subunit is configured to, for each advertisement text in the plurality of advertisement texts, based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, from At least one associated text identified for the advertising text in at least one other advertising text; and The second determination subunit is configured to determine, for each advertisement text among the plurality of advertisement texts, based on the semantics and revenue information of the advertisement text and the corresponding semantics and revenue information corresponding to at least one associated text The feature vector representation corresponding to the advertisement text. 14. The apparatus according to claim 13, wherein the first determining subunit comprises: A construction module configured to construct a text graph containing a plurality of nodes corresponding to the plurality of advertisement texts based on the co-occurrence relationship between the plurality of advertisement texts in the historical advertisement data, wherein, for the Any two advertisement texts of the plurality of texts, in response to determining that the co-occurrence relationship between the two advertisement texts satisfies a preset condition, establishing a connection edge based on nodes corresponding to the two advertisement texts; and The first determination module is configured to, for each node in the plurality of nodes, determine at least one associated node corresponding to the node from at least one other node based on the connection relationship between the plurality of nodes, to At least one associated text corresponding to the corresponding advertisement text is obtained. 15. The device according to claim 14, wherein the second determining subunit comprises: The second determination module is configured to, for each connection edge in at least one connection edge between the node corresponding to the advertisement text and the corresponding at least one associated node, based on the advertisement corresponding to the two endpoints of the connection edge Semantic and yield information of the text, determining the edge vector representation of the connected edge; and The third determining module is configured to determine the feature vector representation of the advertisement text based on the edge vector representation of the at least one connected edge. 16. The apparatus of claim 15, wherein the second determining module is configured to: determining the semantic similarity between the advertisement texts respectively corresponding to the two endpoints; and Based on the semantic similarity and the revenue information of the advertisement text corresponding to the two endpoints, determine the edge vector representation of the connecting edge. 17. The apparatus of claim 16, wherein the second determining module is configured to: Inputting the advertisement texts corresponding to the two endpoints of the connecting edge into the pre-training language model to obtain the semantic similarity output by the pre-training language model, wherein the pre-training language model utilizes labeled corpus data obtained by training. 18. The device according to any one of claims 15-17, wherein the feature vector determining unit further comprises: The third determining subunit is configured to, for each connecting edge in at least one connecting edge between the node corresponding to the advertisement text and the corresponding at least one associated node, determine the corresponding frequency of co-occurrence between ad texts, And wherein, the second determining module is configured to determine the edge vector representation of the connecting edge based on the semantic and revenue information of the advertisement text corresponding to the two endpoints of the connecting edge and the co-occurrence frequency. 19. The device according to claim 18, wherein the eigenvector determining unit further comprises: a processing subunit configured to perform normalization processing on the co-occurrence frequency and the revenue information respectively to obtain a normalized co-occurrence frequency and a normalized revenue information, And wherein, the first determining module is configured to determine the edge of the connecting edge based on the semantics and normalized revenue information of the advertisement text corresponding to the two endpoints of the connecting edge and the normalized co-occurrence frequency Vector representation. 20. The apparatus according to any one of claims 15-19, wherein the second determining module is configured to: inputting the advertisement text corresponding to the two endpoints of the connecting edge and its revenue information into the edge vector encoding model to obtain the edge vector representation output by the edge vector encoding model, Wherein, the edge vector coding model is obtained by training in the following manner: Obtaining a sample text graph containing a plurality of nodes one-to-one corresponding to a plurality of sample texts and revenue information of each sample text in the plurality of sample texts, the sample text graph includes multiple Multiple connection edges of nodes; For each connection edge among the plurality of connection edges included in the sample text graph, input the sample text corresponding to the two end points of the connection edge and its revenue information into the edge vector encoding model to obtain the edge vector an edge vector representation of said connected plate output by the encoding model; Based on the edge vector representations of the plurality of connection edges, determine feature vector representations of a plurality of sample texts corresponding to the plurality of nodes; Acquiring the real correlation between the first sample text and the second sample text among the plurality of sample texts; determining a predicted correlation between the first sample text and the second sample text based on the first feature vector representation and the second feature vector representation respectively corresponding to the first sample text and the second sample text; and Adjusting parameters of the edge vector coding model based on the true correlation and the predicted correlation. 21. The device according to any one of claims 14-20, wherein the plurality of advertisement texts includes a plurality of historical query texts and a plurality of historical recommendation texts, the historical advertisement data includes a plurality of text pairs, the Each of the plurality of text pairs includes a historical query text and historical recommendation text recommended to users based on the historical query text, And where the building blocks are configured as: In response to determining that the historical advertisement data includes a text pair consisting of the two advertisement texts, an edge of the text graph is established with nodes corresponding to the two advertisement texts as vertices. 22. The apparatus according to any one of claims 14-21, wherein the first determining module is configured to: For each of the plurality of other nodes, in response to the number of connection hops between the other node and the node being not greater than a preset threshold, it is determined that the other node is the associated node. 23. The apparatus according to any one of claims 14-21, wherein the first determining module is configured to: Based on the connection relationship between the plurality of nodes and preset rules, determine the respective sampling probabilities of the plurality of other nodes, wherein, according to the preset rules, among the plurality of other nodes and between the nodes The sampling probability of a node with a smaller connection hop is greater than the sampling probability of a node with a larger connection hop between the node; and Randomly sampling the plurality of other nodes based on the sampling probability to obtain the at least one associated node. 24. An advertising text recommendation device, comprising: a second acquiring unit configured to acquire target advertisement text and a plurality of candidate advertisement texts; The apparatus according to any one of claims 13-23, configured to determine a degree of relevance between the target advertisement text and the plurality of candidate advertisement texts; and The second determining unit is configured to determine at least one advertisement text to be recommended from the multiple candidate advertisement texts based on the correlation between the target advertisement text and the multiple candidate advertisement texts. 25. 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, the instructions are executed by the at least one processor, so that the at least one processor can perform any one of claims 1-12 Methods. 26. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method according to any one of claims 1-12. 27. A computer program product comprising a computer program, wherein said computer program, when executed by a processor, implements the method according to any one of claims 1-12.

Images