CN118069248A - Service execution method, device, equipment and readable storage medium - Google Patents
Service execution method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN118069248A CN118069248A CN202410161353.3A CN202410161353A CN118069248A CN 118069248 A CN118069248 A CN 118069248A CN 202410161353 A CN202410161353 A CN 202410161353A CN 118069248 A CN118069248 A CN 118069248A
- Authority
- CN
- China
- Prior art keywords
- node
- behavior sequence
- sequence
- target
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000003860 storage Methods 0.000 title claims abstract description 21
- 230000006399 behavior Effects 0.000 claims description 221
- 238000013519 translation Methods 0.000 claims description 60
- 238000012545 processing Methods 0.000 claims description 24
- 230000000644 propagated effect Effects 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 16
- 230000004044 response Effects 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 12
- 230000000694 effects Effects 0.000 abstract description 8
- 230000004927 fusion Effects 0.000 abstract description 7
- 230000003993 interaction Effects 0.000 abstract description 4
- 230000006870 function Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 11
- 125000004122 cyclic group Chemical group 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 238000012546 transfer Methods 0.000 description 10
- 230000006872 improvement Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000008685 targeting Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 101100481876 Danio rerio pbk gene Proteins 0.000 description 1
- 101100481878 Mus musculus Pbk gene Proteins 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000013058 risk prediction model Methods 0.000 description 1
- 239000010979 ruby Substances 0.000 description 1
- 229910001750 ruby Inorganic materials 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The specification discloses a service execution method, a device, equipment and a readable storage medium, wherein the method, the device and the equipment take each acquired service object as a node to determine a target topological graph, acquire a behavior sequence of each service object, determine a relation weight between the node and a neighboring node of the node according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node aiming at each node in the target topological graph, update the characteristics of the node according to the characteristics of the neighboring node of the node, the behavior sequence of the neighboring node of the node and the relation weight, acquire the target characteristics of the node, and execute the service according to the target characteristics of each node in the target topological graph. Therefore, the information interaction and fusion of the behavior sequence and the graph data are carried out through neighbor relation, propagation feature extraction and node feature update, so that the features of the behavior sequence and the features of the graph data are effectively integrated, and the execution effect of downstream business and the privacy security of the data are improved.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a readable storage medium for executing a service.
Background
With the improvement of people on the attention of private data and the rapid development of artificial intelligence technology, online business is rapidly developed and widely focused. In the actual service, the sequence data (such as the behavior sequence of the user) of the service object and the graph data containing the service object can be respectively analyzed, the characteristics of the service object are obtained through fusion, and the method is applied to the downstream service (such as wind control, personalized recommendation, search and the like), so that the execution effect of the downstream service is improved.
Therefore, how to effectively integrate sequence data and graph data and improve the service execution effect becomes an important problem to be solved.
Disclosure of Invention
The present specification provides a service execution method, apparatus, device, and readable storage medium, to partially solve the above-mentioned problems in the prior art.
The technical scheme adopted in the specification is as follows:
the present specification provides a service execution method, including:
Acquiring each service object and a behavior sequence of each service object, and constructing a target topological graph by taking each service object as a node respectively;
Determining a relation weight between each node and the neighboring node of the node according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node aiming at each node in the target topological graph;
updating the characteristics of the node according to the characteristics of the neighbor nodes of the node, the behavior sequences of the neighbor nodes of the node and the relationship weights to obtain target characteristics of the node;
and executing the service according to the target characteristics of each node in the target topological graph.
The present specification provides a service execution apparatus, including:
The target topological graph determining module is used for acquiring each service object and a behavior sequence of each service object and constructing a target topological graph by taking each service object as a node respectively;
The relation weight determining module is used for determining the relation weight between each node in the target topological graph and the adjacent node of the node according to the behavior sequence of the node and the behavior sequence of the adjacent node of the node;
the feature updating module is used for updating the features of the node according to the features of the neighbor nodes of the node, the behavior sequences of the neighbor nodes of the node and the relation weight to obtain target features of the node;
and the service execution module is used for executing the service according to the target characteristics of each node in the target topological graph.
The present specification provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements the above-described service execution method.
The present specification provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the above-mentioned service execution method when executing the program.
The above-mentioned at least one technical scheme that this specification adopted can reach following beneficial effect:
In the service execution method provided by the specification, a target topological graph is determined by taking each acquired service object as a node, a behavior sequence of each service object is acquired, for each node in the target topological graph, a relation weight between the node and a neighboring node of the node is determined according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node, and the characteristics of the node, the behavior sequence of the neighboring node of the node and the relation weight are updated according to the characteristics of the neighboring node of the node, so that the target characteristics of the node are obtained, and the service is executed according to the target characteristics of each node in the target topological graph. Therefore, the information interaction and fusion of the behavior sequence and the graph data are carried out through the neighbor relation, the propagation feature extraction and the node feature update, so that the more efficient and stable feature extraction and update are carried out based on the multi-mode complementary information of the behavior sequence and the graph data in the process of each feature iteration update, and the execution effect of downstream business is improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the specification, illustrate and explain the exemplary embodiments of the present specification and their description, are not intended to limit the specification unduly. Attached at
In the figure:
fig. 1 is a schematic flow chart of a service execution method in the present specification;
fig. 2 is a schematic flow chart of a service execution method in the present specification;
fig. 3 is a schematic flow chart of a service execution method in the present specification;
Fig. 4 is a schematic flow chart of a service execution method in the present specification;
Fig. 5 is a schematic diagram of a service execution device provided in the present specification;
fig. 6 is a schematic view of the electronic device corresponding to fig. 1 provided in the present specification.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the present specification more apparent, the technical solutions of the present specification will be clearly and completely described below with reference to specific embodiments of the present specification and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present specification. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.
In addition, all the actions for acquiring signals, information or data in the present specification are performed under the condition of conforming to the corresponding data protection rule policy of the place and obtaining the authorization given by the corresponding device owner.
The features of the following examples and embodiments may be combined with each other without any conflict.
Sequence data and graph data are two data types that are more common in practical applications. Sequence data is typically text, sequence of operations, or page access records of business objects, etc., and graph data includes social relationships, business relationships, device interactions, etc. between a plurality of business objects. In an actual service scene, the two data can be obtained at the same time, so that the multi-mode information of the sequence data and the graph data can be effectively integrated, and the execution effect and the execution efficiency of the downstream service can be improved.
In a conventional integration scheme, sequence features are generally extracted from sequence data of a service object, and simultaneously, information extraction is performed on graph data containing the service object in parallel, and the sequence features and information of the service object extracted from the graph data are fused to obtain fusion features. Or taking the graph data as elements of the sequence, constructing the sequence containing a plurality of graph data, extracting and characterizing each element in the sequence, namely the graph data, and extracting sequence information by using a sequence feature extraction model. Or firstly obtaining the sequence features of the business object through the sequence feature extraction model, and then correlating the sequence features to the graph for feature propagation.
In the scheme, the information of the other mode cannot be referred to in the process of extracting the characteristics of the data of the certain mode, so that the characteristics are lost or repeated.
Based on the above, the present specification provides a service execution method, which performs information interaction and fusion of a behavior sequence and graph data in a neighbor relation, propagation feature extraction and node feature update, so as to effectively integrate features of the behavior sequence and features of the graph data, thereby improving the execution effect of downstream services.
The following describes in detail the technical solutions provided by the embodiments of the present specification with reference to the accompanying drawings.
Fig. 1 is a schematic flow chart of a service execution method provided in the present specification.
S100: and obtaining each service object and a behavior sequence of each service object, and constructing a target topological graph by taking each service object as a node.
In the embodiments of the present specification, a service execution method is provided, which can be executed by an electronic device such as a server for executing a service.
In the computer technology, the topological graph can be used for representing the relation among the nodes, the information of the neighbor nodes of each node is aggregated through feature aggregation, and the features of the nodes are determined by combining the information of the nodes, so that the topological graph can be applied to various business scenes such as user wind control, personalized recommendation, information search, advertisement orientation and the like. For example, a funding flow relationship between accounts, a device sharing relationship between accounts, a relationship between owners of accounts, and the like can be represented by a topology map.
In this embodiment of the present disclosure, a plurality of nodes in the target topology may be used to represent a plurality of service objects respectively, different nodes correspond to different service objects, a service object may be a user or an account of the user, and edges between the nodes are used to represent that a service relationship exists between the service objects, such as a fund transaction relationship between the accounts, and a common relationship of operation devices between the accounts. In the embodiment of the present disclosure, a specific technical scheme is described by taking a service object as an account of a user and a service relationship as a fund business relationship as an example.
In practical application, the characteristics and the behavior sequence of the business object corresponding to each node in the target topological graph can be obtained and used as the characteristics and the behavior sequence corresponding to the node.
The characteristics corresponding to each node can be determined according to the information of the service object corresponding to each node, and the information of the service object can be attribute information of the service object itself or data of the service related to the service object. For example, in the case where the business object is an account, the information of the business object may be information such as an account opening line of the account, a type of electronic device using the account, identity information of an account holder, a registration place of the account, and the like. The method for determining the characteristics of the node according to the information of the service object may be any existing method such as manually constructing a structured characteristic, extracting the characteristic through a pre-trained neural network, and the like, which is not limited in this specification.
The behavior sequence of each node is actually a sequence obtained by sequencing a plurality of behaviors of a corresponding business object in a certain period. The behavior of a business object is generally associated with the type of business object, e.g., where the business object is an account, the behavior of the business object may be to transfer funds, and then the sequence of behavior of the business object may be the amount of funds transferred multiple times over a period of time, such as [ transfer funds 1000, transfer funds 2000, transfer funds 3000, transfer funds 4000]. Of course, the behaviors corresponding to the elements in the behavior sequence of the node may be the same or different, and this description is not limited to this. Still as an example, the behavior of the business object is to transfer funds, and the behavior sequence of the business object may be the amount of money that is transferred in and out during a certain period of time, such as [ transfer funds 1000, transfer funds 2000, transfer funds 3000, transfer funds 4000].
Optionally, in this specification, with the target topology diagram as G, in the target topology diagram G, N nodes v= { V n |n=1, … … N } and M edges e= { E m＝(vi,vj)|m＝1,……M;vi,vj E V } are used, where each node V n＝{hn,Sn } is characterized by h n and the behavior sequence is S n, where S n [ t ], t=1, … …, and L.
S102: and aiming at each node in the target topological graph, determining the relation weight between the node and the neighbor node of the node according to the behavior sequence of the node and the behavior sequence of the neighbor node of the node.
Specifically, in the graph model, for each node, the characteristics of the node are typically iteratively updated by a propagated aggregate calculation of the information of the neighboring nodes. Therefore, it is first necessary to determine the relationship weight between each node and the corresponding neighbor node in the target topology graph, so as to measure the amount of information that the information of the neighbor node propagates to the node.
Specifically, an alternative implementation of a conventional relationship weight that measures the relationship strength between a node and all its neighboring nodes may be as follows:
in the above formula, the relationship weights are actually determined based on the characteristics of the node and the characteristics of all its neighbor nodes.
However, in this specification, since the behavior sequence of each node is introduced into the target topology graph, the relationship weight between the node and the neighboring node may be defined only based on the correlation between the behavior sequence of the node and the behavior sequence of the neighboring node, and an alternative implementation manner is as follows:
wherein, Is the correlation between the behavior sequence S i of the node v i and the behavior sequence S j of the neighboring node v j.
Of course, in order to keep the association relation of the information between the node and the neighbor node as far as possible, the relation weight can be determined by introducing the behavior sequence of the node and the behavior sequences of all the neighbor nodes on the basis of determining the relation weight by the correlation of the characteristics of the node and the characteristics of the neighbor node. The following is an alternative implementation:
wherein, Is the correlation between the behavior sequence S i of the node v i and the behavior sequence S j of the neighboring node v j.
In this specification, the correlation between the behavior sequence of the node and the behavior sequence of the neighboring node of the node may be determined based on the similarity between each element in the behavior sequence of the node and each element in the corresponding position in the behavior sequence of the neighboring node of the node, that is, the behavior sequence of the node and the behavior sequence of the neighboring node of the node are dot-product. An alternative implementation is as follows:
In addition, in the present specification, one or more neighboring nodes corresponding to each node are determined by edges in the target topology graph, and the neighboring nodes directly connected to each node by the edges are actually one-hop neighboring nodes of the node, which is not limited in the number of neighboring nodes of each node and the hop count of the neighboring nodes to the node in the present specification.
S104: and updating the characteristics of the node according to the characteristics of the neighbor nodes of the node, the behavior sequences of the neighbor nodes of the node and the relationship weights to obtain the target characteristics of the node.
Specifically, in the conventional graph model, information that needs to be propagated to a node is generally determined based on information of a neighboring node, that is, features of the node are updated based on features of the neighboring node. However, in the present specification, since each node in the target topology graph has a feature and a behavior sequence, information propagated to the node may be determined according to the feature of the neighboring node of the node and the behavior sequence of the neighboring node of the node, and then, the feature of the node is updated by combining with the relationship weight, so as to obtain the target feature of the node.
Specifically, the first step: and extracting the behavior sequence characteristics of the neighbor nodes of the node from the behavior sequences of the neighbor nodes of the node.
In this step, a sequence extraction model may be used to perform feature extraction on the behavior sequence of the neighboring node, to obtain the behavior sequence feature of the neighboring node, and the behavior sequence feature of the neighboring node is used as a part of the information transmitted to the node. In the present specification, the sequence extraction model for extracting the behavior sequence features is not limited, and may be any existing model, such as a transducer.
And a second step of: and determining the feature to be propagated according to the feature of the neighbor node of the node and the behavior sequence feature of the neighbor node of the node.
And then, the characteristics of the neighbor nodes of the node are also used as a part of information transmitted to the node, and are spliced with the behavior sequence characteristics of the neighbor nodes of the node, and an alternative implementation mode is as follows:
Where transducer () is a sequence feature extraction model and linear () is a linear layer function.
And a third step of: and updating the characteristics of the node according to the relation weight and the characteristics to be propagated to obtain the target characteristics of the node.
Further, based on the relationship weights, the influence of the feature to be propagated on the node features is determined. Generally, the relation weight is taken as the weight, and the characteristics to be propagated are weighted, so that the update of the neighbor node of the node to the characteristics of the node can be determined.
The following is an alternative implementation:
And then, according to the characteristics of the node and the weighted characteristics to be propagated, the target characteristics of the node can be obtained, wherein the target characteristics of the node comprise information of the node, information of a behavior sequence of a neighbor node of the node and characteristic information, the characteristics and the behavior sequence of the node are fully integrated, and multi-mode information fusion of the behavior sequence and the characteristics is carried out in the characteristics propagation and the state updating processes, so that more effective information is extracted for downstream service use, and the execution effect and the execution efficiency of the downstream service are improved.
In addition, it should be noted that, the target features of each node in the target topology graph may be propagated through multiple rounds of iterative update, and the steps S102 to S104 may be a one-time propagation process, so that multiple iterative propagation may be performed in practical application, and the rich information of the neighboring nodes may be propagated to the nodes as far as possible, so as to more effectively integrate the behavior sequences and the multi-modal information of the features. In the present specification, the target feature of the node may be obtained based on one propagation, or may be obtained based on a plurality of propagation, which is not limited in the present specification.
S106: and executing the service according to the target characteristics of each node in the target topological graph.
In the present specification, the services executed according to the target features of each node in the target topology graph include a wind control service for identifying the risk of the service object corresponding to the node, a personalized recommendation service for recommending goods or information for the service object corresponding to the node, an information search service for returning a search result to the service object corresponding to the node, or an advertisement targeting service for targeting a propagation scene for the service object corresponding to the node. In general, all business scenarios that need to be predicted based on abstract feature representations can be executed based on the target features of each node in the target topology graph in the present specification.
In practical applications, downstream traffic may be performed based on target characteristics of each node in the target topology. Specifically, target characteristics of designated nodes in a target topological graph are input into a prediction model, and a prediction result corresponding to the designated nodes of the output of the prediction model is obtained, so that a service is executed based on the prediction result.
For example, in a wind control scenario, the business object corresponding to the designated node may be an account to be detected with risk, and the prediction result obtained based on the pre-trained risk prediction model is actually a risk prediction result, where the risk prediction result is used to indicate whether the account to be detected has risk, so if the risk prediction result indicates that the account to be detected has risk, freezing, transaction allowance, real-time transaction monitoring and the like may be performed on the account to be detected. In another example, in the personalized recommendation scenario, the service object corresponding to the designated node may be a consumer, the prediction result obtained by the pre-trained recommendation model based on the target feature of the designated node may be a commodity that the consumer may be interested in, and the commodity that the consumer may be interested in is recommended to the consumer, so that the personalized recommendation service is realized.
In the service execution method provided by the description, a target topological graph is determined by taking each acquired service object as a node, a behavior sequence of each service object is acquired, for each node in the target topological graph, a relation weight between the node and a neighboring node of the node is determined according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node, the characteristics of the node are updated according to the characteristics of the neighboring node of the node, the behavior sequence of the neighboring node of the node and the relation weight, the target characteristics of the node are obtained, and the service is executed according to the target characteristics of each node in the target topological graph.
Therefore, the information interaction and fusion of the behavior sequence and the graph data are carried out through neighbor relation, propagation feature extraction and node feature update, so that the features of the behavior sequence and the features of the graph data are effectively integrated, and the execution effect of downstream business is improved.
In one or more embodiments of the present disclosure, the relationship weight between the node and the neighboring node in the S102 target topology map may be determined based on the sequence, or may be determined based on the sequence in combination with the relationship between the feature of the node and the feature of the neighboring node, which is specifically as follows, as shown in fig. 2:
S200: and determining the sequence correlation of the node and the neighbor node of the node according to the similarity between the behavior sequence of the node and the behavior sequence of the neighbor node of the node.
As previously indicated, the sequence correlation between the behavior sequence of the node and the behavior sequence of the neighboring node may be determined based on a similarity between the behavior sequence of the node and the behavior sequence of the neighboring node of the node, where the similarity may be determined by a dot product of the behavior sequence of the node and the behavior sequence of the neighboring node.
S202: and determining the similarity between the characteristics of the node and the characteristics of the neighbor nodes of the node, and taking the similarity as the characteristic correlation of the node and the neighbor nodes of the node.
In this step, the feature of the node and the feature of the neighboring node of the node are vectors with the same dimension, so the similarity between the feature of the node and the feature of the neighboring node of the node may be obtained by using any existing vector similarity determining method, such as pre-similarity, euclidean distance, manhattan distance, pearson correlation coefficient, and the like, which is not limited in this specification.
In general, the greater the similarity between the features of the node and the features of the neighboring nodes of the node, the higher the feature correlation of the node and the neighboring nodes of the node, and the greater the probability that the features of the neighboring nodes propagate to the node.
S204: and determining the relation weight between the node and the neighbor node of the node according to the sequence correlation and the characteristic correlation.
Specifically, the sequence correlation and the feature correlation may be summed to obtain a relationship weight between the node and the neighboring node of the node. The relation weight is used for representing the intensity of the business relation between the node and the neighbor node of the node, and simultaneously, the information quantity of the neighbor node of the node, which is transmitted to the node, is represented, and the larger the relation weight is, the larger the information quantity of the neighbor node (the characteristic and/or the behavior sequence) which is transmitted to the node in the characteristic transmission process is, otherwise, the smaller the relation weight is, the smaller the information quantity of the neighbor node (the characteristic and/or the behavior sequence) which is transmitted to the node in the characteristic transmission process is.
Further, in one or more embodiments of the present disclosure, when the S200 determines a sequence correlation between a behavior sequence of a node and a behavior sequence of a neighboring node, a cyclic right shift process may be further performed on each element in the behavior sequence of the neighboring node, so as to determine the sequence correlation based on a similarity between the processed behavior sequence of the neighboring node and the behavior sequence of the node. This is because the elements in the behavior sequence are generally arranged based on the occurrence order of the behaviors corresponding to the elements, and the behavior sequence of the node and the behavior sequence of the neighboring node may have a temporal offset, so that similar behaviors occur in different orders. If sequence correlation is determined based on only the similarity of corresponding elements of the behavior sequences of the nodes and the neighboring nodes, there may be a lower sequence correlation between two behavior sequences that are similar in behavior but differ in chronological order, thereby losing information of part of the behavior sequences. Based on this, the following scheme is further implemented for S200 in the present specification, as shown in fig. 3:
S300: and (3) performing iteration: determining a current translation distance according to a last translation distance, performing circular right shift processing on a behavior sequence of a neighboring node of the node according to the current translation distance to obtain a processed behavior sequence of the neighboring node of the node, and determining a current sequence correlation between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node according to the similarity between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node until each translation distance is traversed, wherein the current translation distance is greater than the last translation distance.
Specifically, as described in S102, the similarity determined by the dot product of the behavior sequence of the node and the behavior sequence of the neighboring node of the node is actually strictly determined according to the occurrence sequence of each behavior of the service object corresponding to the node and the occurrence sequence of each behavior of the service object corresponding to the neighboring node of the node, so as to determine whether the behaviors of the two service objects are similar at each occurrence moment of the behavior in turn. However, in practical applications, there may be a problem that the two business objects behave similarly, but the occurrence time has an offset.
An alternative implementation is: the behavior sequence of the node a is S A = {0, a, B, c }, and the behavior sequence of the neighboring node B of the node a is S B = { a, B, c, 0}, wherein a, B, c represent three different behaviors respectively. If the similarity of the node A and the node B is determined based on the dot product between the behavior sequences, the similarity is 0, which indicates that the node A and the node B are dissimilar. However, the node a and the node B do take the same actions, and only the occurrence time is different, so that the node a and the node B are not similar based on dot product between action sequences, and the problem of information loss can exist. To address the above problem, a round-robin right-shift may be performed for each element in the behavior sequence of the node B. When the translation distance is 3, the behavior sequence of the node B after the cyclic right shift processing is S' B = {0, a, B, c }. Obviously, the behavior sequence of the processed node B is completely the same as that of the node A, so that the similarity between the behavior sequence of the processed node B and that of the node A is 1, and the behavior sequence of the processed node B is completely different from that of the node A before processing.
Therefore, the sequence correlation between the behavior sequence of the node with similar behaviors but misaligned behavior occurrence time and the behavior sequence of the neighbor node can be accurately obtained by adopting the cyclic right shift processing to the neighbor node.
In this specification, before determining the sequence correlation, it is unclear whether the above-mentioned problem exists between the behavior sequence of the node and the behavior sequence of the neighboring node of the node, so that a plurality of translation distances can be determined according to the length of the behavior sequence, and each translation distance is traversed to obtain the sequence correlation between the behavior sequence of the node and the behavior sequence of the neighboring node of the node after the cycle right shift based on each translation distance. It will be appreciated that the length of the sequence of behavior of each node in the target topology is the same.
To this end, the following scheme may be performed iteratively:
First, a current translation distance is determined based on a last translation distance. Wherein the current translation distance is greater than the last translation distance. In general, the size of the difference between the translation distances of two adjacent iterations (the difference between the current translation distance and the last translation distance) can be predetermined according to the actual scene, and the number of translation distances is determined by the length of the behavior sequence of the node. For convenience of description, in this specification, a specific scheme is described by taking a difference of translation distances between two adjacent iterations as1, a translation distance of a first iteration as 0, and a behavior sequence length of the node as L as an example. It is known that the translation distance τ is in the range of [0, L ]. And, in this step, the current translation distance is determined based on the difference between the last translation distance and the translation distance of the two adjacent iterations. And, the current translation distance is greater than the last translation distance. For example, when the one-time translation distance is 1, the current translation distance is 2.
And secondly, performing circular right shift processing on the behavior sequence of the neighbor node of the node according to the current translation distance to obtain the processed behavior sequence of the neighbor node of the node. Specifically, the cyclic right shift is to shift each element in the behavior sequence of the neighboring node of the node rightward by the current translation distance unit, and the shifted low order is placed at the high order of the behavior sequence. The following is an alternative implementation: the behavior sequence of the neighbor node of the node before the cyclic right shift processing is S j[t]＝{s1,s2,s3,……,sL-2,sL-1,sL, and the current translation distance is 2, and the behavior sequence of the neighbor node of the node after the cyclic right shift processing is S j[t-2]＝{sL-1,sL,s1,s2,s3,……,sL-2.
Then, according to the similarity between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node, determining the current sequence correlation between the processed behavior sequences of the node and the neighboring node of the node.
This step still uses dot product to determine the sequence correlation. Specifically, according to the summation of dot products between each element in the behavior sequence of the node and each element in the corresponding position in the processed behavior sequence of the neighboring node of the node, determining the current sequence correlation between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node.
An alternative implementation is as follows:
Wherein L is the length of the behavior sequence of the node, S i [ t ] represents the behavior sequence of the node, and S j [ t-tau ] represents the behavior sequence of the neighbor node of the node after cyclic right-shift processing with the current translation distance as tau.
And finally, traversing each translation distance, and circularly executing the scheme to obtain a plurality of current sequence correlations obtained by traversing the translation distance tau from 0 to L.
S302: a preset number of specified sequence correlations is selected from each current sequence correlation.
In the present specification, the current sequence correlations obtained based on S300 are sequentially arranged from large to small, and a predetermined number of specified sequence correlations are selected based on the actual application scenario. The specific manner of selecting the specified sequence correlation is not limited in this specification, and a larger preset number of the current sequence correlations may be selected as the specified sequence correlation, or a smaller preset number of the current sequence correlations may be selected as the specified sequence correlation.
Optionally, a larger preset number of current sequence correlations is selected from the current sequence correlations as the specified sequence correlations, that is, topK is taken, where K is the preset number, and may be manually determined in advance based on the actual application scenario.
An alternative implementation is as follows:
Wherein, T (v i,vj) represents K larger appointed sequence relatives selected from each current sequence relativity, and each appointed sequence relativity corresponds to a translation distance tau 1,…,τK respectively.
S304: and determining the sequence correlation of the node and the neighbor nodes of the node according to the preset number of specified sequence correlations.
Specifically, adding the preset number of specified sequence correlations to obtain the sequence correlation between the node and the neighboring node of the node.
In addition, the specified sequence correlations may also be normalized prior to summation, an alternative embodiment as follows:
based on the specified sequence correlation after normalization, an alternative implementation mode of the sequence correlation between the node and the neighbor node of the node is as follows:
Thus, based on the scheme shown in fig. 3, in the foregoing S102, an alternative implementation of the relationship weights is as follows:
In one or more embodiments of the present disclosure, the foregoing solutions are all propagation iterative updating for the feature of each node in the target topology graph, and the behavior sequence of the node is still static and not updated, but in order to further integrate the information of the feature and the behavior sequence, the behavior sequence of the node may also be propagated and updated based on the feature of the neighboring node and the behavior sequence of the neighboring node. Therefore, in the present specification, before S106 shown in fig. 1, the behavior sequence of the node may be updated based on the elements in the behavior sequence of the neighboring node, so as to implement iterative propagation and update of the behavior sequence of the node in the target topology. The specific scheme is as follows, as shown in fig. 4:
In fig. 3, element similarity between each element in the behavior sequence of the node and each element in the behavior sequence of the neighboring node is determined, and a preset number of specified element similarities are determined based on the element similarities. The specified element similarity is used for indicating that the similarity between a first element in the behavior sequence of the node for determining the specified element similarity and a second element in the behavior sequence of the neighbor node is higher.
S400: and determining a reference sequence according to the characteristics of the neighbor nodes of the node and each element contained in the behavior sequence of the neighbor nodes of the node.
Specifically, the characteristics of the neighbor node of the node are respectively spliced with each element in the behavior sequence of the neighbor node of the node, so that a reference sequence with the same length as the behavior sequence of the neighbor node of the node and overlapping the characteristics of the neighbor node of the node is obtained.
An alternative implementation is as follows:
where linear () is a linear layer function.
S402: determining the preset number of specified translation distances according to the correlation of the preset number of specified sequences, and respectively performing circular right shift processing on the reference sequences according to the preset number of specified translation distances to obtain a plurality of processed reference sequences.
Specifically, a preset number of specified sequence correlations are determined in the scheme shown in fig. 3, where the specified sequence correlations indicate that the behavior sequence of the node has a higher similarity to the behavior sequence of the neighboring node of the node after the cyclic right shift processing, that is, after the cyclic right shift processing is performed on the neighboring node of the node by using the translation distance for determining the specified sequence correlations, the relationship between the behavior sequence of the node and the behavior sequence of the neighboring node of the node after the processing is stronger. Thus, the behavior sequence of the neighbor node of the node after the processing with the stronger relationship can be used as a part of the information of the behavior sequence propagated to the node. And another part of the information propagated to the behavior sequence of the node is characteristic of the neighboring nodes of the node.
Thus, the preset number of specified translation distances respectively adopted for obtaining the preset number of specified sequence correlations may be determined based on the preset number of specified sequence correlations.
And then, for each specified translation distance, performing circular right shift processing on the reference sequence according to the specified translation distance to obtain a processed reference sequence.
In this way, the reference sequence may be aligned with the specified sequence correlation determined by the specified translation distance, thereby facilitating weighting in subsequent steps.
An alternative implementation is as follows:
Where Roll () is a function of the cyclic right shift of the reference sequence Message S(vj), τ k is the specified translation distance.
S404: and updating the behavior sequence of the node according to the preset number of specified sequence correlations and the plurality of processed reference sequences to obtain a target behavior sequence of the node.
Further, with the correlation of the preset number of specified sequences as weights, weighting the plurality of processed reference sequences obtained based on the preset number of specified translation distances, and then summing to obtain updated information of the behavior sequence of the neighboring node of the node to the node, an optional implementation manner is as follows:
And further, based on the updated information of the behavior sequence of the neighbor node of the node to the node and the behavior sequence of the node, the updated target behavior sequence of the node can be obtained.
Thus, based on the scheme shown in fig. 4, S106 may be configured to execute a service according to the target feature of each node in the target topology map and the target behavior sequence of each node.
Specifically, first, in response to a service execution request, a designated node is determined from the nodes included in the target topology.
And then, taking the target characteristics and the target behavior sequence of the designated node as input, and inputting the target characteristics and the target behavior sequence into a pre-trained prediction model to obtain a prediction result corresponding to the designated node, which is output by the prediction model.
And finally, executing the service according to the prediction result corresponding to the designated node.
The above service execution method provided for one or more embodiments of the present disclosure further provides a corresponding service execution device based on the same concept, as shown in fig. 5.
Fig. 5 is a schematic diagram of a service execution device provided in the present specification, which specifically includes:
The target topological graph determining module 500 is configured to obtain each service object and a behavior sequence of each service object, and construct a target topological graph by using each service object as a node;
a relationship weight determining module 502, configured to determine, for each node in the target topology graph, a relationship weight between the node and a neighboring node of the node according to a behavior sequence of the node and a behavior sequence of the neighboring node of the node;
A feature updating module 504, configured to update the feature of the node according to the feature of the neighboring node of the node, the behavior sequence of the neighboring node of the node, and the relationship weight, to obtain a target feature of the node;
And the service execution module 506 is configured to execute a service according to the target characteristics of each node in the target topology graph.
Optionally, the relationship weight determining module 502 is specifically configured to determine, according to a similarity between the behavior sequence of the node and the behavior sequence of the neighboring node of the node, a sequence correlation between the node and the neighboring node of the node; determining the similarity between the characteristics of the node and the characteristics of the neighbor nodes of the node, and taking the similarity as the characteristic correlation between the node and the neighbor nodes of the node; and determining the relation weight between the node and the neighbor node of the node according to the sequence correlation and the characteristic correlation.
Optionally, the relation weight determining module 502 is specifically configured to iteratively perform: determining a current translation distance according to a last translation distance, performing circular right shift processing on a behavior sequence of a neighboring node of the node according to the current translation distance to obtain a processed behavior sequence of the neighboring node of the node, and determining a current sequence correlation between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node according to the similarity between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node until each translation distance is traversed, wherein the current translation distance is greater than the last translation distance; selecting a preset number of specified sequence correlations from each current sequence correlation; and determining the sequence correlation of the node and the neighbor nodes of the node according to the preset number of specified sequence correlations.
Optionally, the feature updating module 504 is specifically configured to extract, from a behavior sequence of a neighboring node of the node, a behavior sequence feature of the neighboring node of the node; determining a feature to be propagated according to the feature of the neighbor node of the node and the behavior sequence feature of the neighbor node of the node; and updating the characteristics of the node according to the relation weight and the characteristics to be propagated to obtain the target characteristics of the node.
Optionally, the apparatus further comprises:
The sequence updating module 508 is specifically configured to determine a reference sequence according to characteristics of a neighboring node of the node and elements included in a behavior sequence of the neighboring node of the node; determining the preset number of specified translation distances according to the correlation of the preset number of specified sequences, and respectively performing circular right shift processing on the reference sequences according to the preset number of specified translation distances to obtain a plurality of processed reference sequences; updating the behavior sequence of the node according to the preset number of specified sequence correlations and the plurality of processed reference sequences to obtain a target behavior sequence of the node;
Optionally, the service execution module 506 is specifically configured to execute a service according to the target feature of each node in the target topology graph and the target behavior sequence of each node.
Optionally, the service includes one of a wind control service, a recommendation service, and a search service.
The present specification also provides a computer-readable storage medium storing a computer program operable to execute the service execution method shown in fig. 1 described above.
The present specification also provides a schematic structural diagram of the electronic device shown in fig. 6. At the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile storage, as illustrated in fig. 6, although other hardware required by other services may be included. The processor reads the corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to implement the service execution method shown in fig. 1. Of course, other implementations, such as logic devices or combinations of hardware and software, are not excluded from the present description, that is, the execution subject of the following processing flows is not limited to each logic unit, but may be hardware or logic devices.
In the 90 s of the 20 th century, improvements to one technology could clearly be distinguished as improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) or software (improvements to the process flow). However, with the development of technology, many improvements of the current method flows can be regarded as direct improvements of hardware circuit structures. Designers almost always obtain corresponding hardware circuit structures by programming improved method flows into hardware circuits. Therefore, an improvement of a method flow cannot be said to be realized by a hardware entity module. For example, a programmable logic device (Programmable Logic Device, PLD) (e.g., field programmable gate array (Field Programmable GATE ARRAY, FPGA)) is an integrated circuit whose logic functions are determined by user programming of the device. A designer programs to "integrate" a digital system onto a PLD without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Moreover, nowadays, instead of manually manufacturing integrated circuit chips, such programming is mostly implemented with "logic compiler (logic compiler)" software, which is similar to the software compiler used in program development and writing, and the original code before being compiled is also written in a specific programming language, which is called hardware description language (Hardware Description Language, HDL), but HDL is not just one, but a plurality of kinds, such as ABEL(Advanced Boolean Expression Language)、AHDL(Altera Hardware Description Language)、Confluence、CUPL(Cornell University Programming Language)、HDCal、JHDL(Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby Hardware Description Language), and VHDL (Very-High-SPEED INTEGRATED Circuit Hardware Description Language) and Verilog are currently most commonly used. It will also be apparent to those skilled in the art that a hardware circuit implementing the logic method flow can be readily obtained by merely slightly programming the method flow into an integrated circuit using several of the hardware description languages described above.
The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, application SPECIFIC INTEGRATED Circuits (ASICs), programmable logic controllers, and embedded microcontrollers, examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic of the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller may thus be regarded as a kind of hardware component, and means for performing various functions included therein may also be regarded as structures within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.
The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer chip or entity, or by a product having a certain function. One typical implementation is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
For convenience of description, the above devices are described as being functionally divided into various units, respectively. Of course, the functions of each element may be implemented in one or more software and/or hardware elements when implemented in the present specification.
It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present description is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the specification. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.
The foregoing is merely exemplary of the present disclosure and is not intended to limit the disclosure. Various modifications and alterations to this specification will become apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, or the like, which are within the spirit and principles of the present description, are intended to be included within the scope of the claims of the present description.
Claims (10)
1. A method of service execution, the method comprising:
Acquiring each service object and a behavior sequence of each service object, and constructing a target topological graph by taking each service object as a node respectively;
Determining a relation weight between each node and the neighboring node of the node according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node aiming at each node in the target topological graph;
updating the characteristics of the node according to the characteristics of the neighbor nodes of the node, the behavior sequences of the neighbor nodes of the node and the relationship weights to obtain target characteristics of the node;
and executing the service according to the target characteristics of each node in the target topological graph.
2. The method according to claim 1, wherein the determining the relation weight between the node and the neighboring node of the node according to the behavior sequence of the node and the behavior sequence of the neighboring node of the node specifically comprises:
Determining the sequence correlation of the node and the neighbor node of the node according to the similarity between the behavior sequence of the node and the behavior sequence of the neighbor node of the node;
Determining the similarity between the characteristics of the node and the characteristics of the neighbor nodes of the node, and taking the similarity as the characteristic correlation between the node and the neighbor nodes of the node;
And determining the relation weight between the node and the neighbor node of the node according to the sequence correlation and the characteristic correlation.
3. The method according to claim 2, wherein the determining the sequence correlation between the node and the neighboring node of the node according to the similarity between the behavior sequence of the node and the behavior sequence of the neighboring node of the node specifically comprises:
And (3) performing iteration: determining a current translation distance according to a last translation distance, performing circular right shift processing on a behavior sequence of a neighboring node of the node according to the current translation distance to obtain a processed behavior sequence of the neighboring node of the node, and determining a current sequence correlation between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node according to the similarity between the behavior sequence of the node and the processed behavior sequence of the neighboring node of the node until each translation distance is traversed, wherein the current translation distance is greater than the last translation distance;
selecting a preset number of specified sequence correlations from each current sequence correlation;
and determining the sequence correlation of the node and the neighbor nodes of the node according to the preset number of specified sequence correlations.
4. The method of claim 1, wherein the updating the characteristic of the node according to the characteristic of the neighboring node of the node, the behavior sequence of the neighboring node of the node, and the relationship weight to obtain the target characteristic of the node specifically comprises:
Extracting the behavior sequence characteristics of the neighbor nodes of the node from the behavior sequences of the neighbor nodes of the node;
determining a feature to be propagated according to the feature of the neighbor node of the node and the behavior sequence feature of the neighbor node of the node;
And updating the characteristics of the node according to the relation weight and the characteristics to be propagated to obtain the target characteristics of the node.
5. The method of claim 3, further comprising, prior to said executing the traffic according to the target characteristics of each node in the target topology,:
Determining a reference sequence according to the characteristics of the neighbor nodes of the node and each element contained in the behavior sequence of the neighbor nodes of the node;
Determining the preset number of specified translation distances according to the correlation of the preset number of specified sequences, and respectively performing circular right shift processing on the reference sequences according to the preset number of specified translation distances to obtain a plurality of processed reference sequences;
Updating the behavior sequence of the node according to the preset number of specified sequence correlations and the plurality of processed reference sequences to obtain a target behavior sequence of the node;
The executing service according to the target characteristics of each node in the target topological graph specifically comprises the following steps:
And executing the service according to the target characteristics of each node in the target topological graph and the target behavior sequence of each node.
6. The method of claim 5, wherein the executing the service according to the target feature of each node in the target topology graph and the target behavior sequence of each node specifically comprises:
determining a designated node from all nodes contained in the target topological graph in response to a service execution request;
taking the target characteristics and the target behavior sequence of the designated node as input, and inputting the target characteristics and the target behavior sequence into a pre-trained prediction model to obtain a prediction result corresponding to the designated node output by the prediction model;
and executing the service according to the prediction result corresponding to the designated node.
7. The method of claim 1, the service comprising one of a wind control service, a recommendation service, a search service.
8. A service execution apparatus comprising:
The target topological graph determining module is used for acquiring each service object and a behavior sequence of each service object and constructing a target topological graph by taking each service object as a node respectively;
The relation weight determining module is used for determining the relation weight between each node in the target topological graph and the adjacent node of the node according to the behavior sequence of the node and the behavior sequence of the adjacent node of the node;
the feature updating module is used for updating the features of the node according to the features of the neighbor nodes of the node, the behavior sequences of the neighbor nodes of the node and the relation weight to obtain target features of the node;
and the service execution module is used for executing the service according to the target characteristics of each node in the target topological graph.
9. A computer readable storage medium storing a computer program which, when executed by a processor, implements the method of any of the preceding claims 1-7.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any of the preceding claims 1-7 when the program is executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410161353.3A CN118069248A (en) | 2024-02-04 | 2024-02-04 | Service execution method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410161353.3A CN118069248A (en) | 2024-02-04 | 2024-02-04 | Service execution method, device, equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118069248A true CN118069248A (en) | 2024-05-24 |
Family
ID=91103460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410161353.3A Pending CN118069248A (en) | 2024-02-04 | 2024-02-04 | Service execution method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118069248A (en) |
-
2024
- 2024-02-04 CN CN202410161353.3A patent/CN118069248A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11494204B2 (en) | Mixed-grained detection and analysis of user life events for context understanding | |
| CN110119860B (en) | Rubbish account detection method, device and equipment | |
| CN110008991B (en) | Risk event identification method, risk identification model generation method, risk event identification device, risk identification equipment and risk identification medium | |
| US20180189356A1 (en) | Detection and analysis of user life events in a communication ecosystem | |
| CN111401766B (en) | Model, service processing method, device and equipment | |
| CN113688313A (en) | Training method of prediction model, information pushing method and device | |
| CN113516480A (en) | Payment risk identification method, device and equipment | |
| CN109428900B (en) | Data processing method and device | |
| CN116049761A (en) | Data processing method, device and equipment | |
| CN116069933A (en) | Business wind control method and device, storage medium and electronic equipment | |
| CN109376988B (en) | Service data processing method and device | |
| CN112836829B (en) | Personalized privacy protection learning method, device and equipment | |
| CN110264213A (en) | A kind of processing method of information, device and equipment | |
| CN116757278B (en) | Training method and device of prediction model, storage medium and electronic equipment | |
| CN117235530A (en) | Method and device for training intention prediction model and electronic equipment | |
| CN118069248A (en) | Service execution method, device, equipment and readable storage medium | |
| CN115545943A (en) | Map processing method, device and equipment | |
| CN115204395A (en) | Data processing method, device and equipment | |
| CN114331602A (en) | Model training method based on transfer learning, information recommendation method and device | |
| CN114638998A (en) | Model updating method, device, system and equipment | |
| CN113205377A (en) | Information recommendation method and device | |
| CN113343085A (en) | Information recommendation method and device, storage medium and electronic equipment | |
| CN117909926A (en) | Risk identification method and device, storage medium and electronic equipment | |
| CN117436512A (en) | Model training method and device | |
| CN115827880B (en) | Business execution method and device based on emotion classification |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |