CN113344369A

CN113344369A - Method and device for attributing image data, electronic equipment and storage medium

Info

Publication number: CN113344369A
Application number: CN202110605781.7A
Authority: CN
Inventors: 张笑雪; 李曼丽; 文晋京; 胡屹
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2021-05-31
Filing date: 2021-05-31
Publication date: 2021-09-03

Abstract

The utility model provides an attribution method of portrait data, which relates to the field of artificial intelligence and can be applied to the field of financial science and technology, comprising the following steps: acquiring marketing portrait data; calculating an evaluation index score of the marketing portrait according to the marketing portrait data; acquiring behavior data corresponding to the marketing picture with the evaluation index score larger than a preset threshold value; and constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and extracting behavior characteristics influencing the evaluation index score in the behavior data. The present disclosure also provides an attribution device of portrait data, an electronic device and a computer readable storage medium.

Description

Method and device for attributing image data, electronic equipment and storage medium

Technical Field

The disclosure relates to the field of artificial intelligence, can be applied to the field of financial science and technology, and particularly relates to an image data attribution method and device, electronic equipment and a storage medium.

Background

In the field of financial services, the portrait data is beneficial to the financial enterprise to comprehensively evaluate and attribute related businesses, so that business optimization suggestions are provided, and the overall business level and capacity of the financial enterprise are improved. For example, for a bank, a client manager and a wealth counselor are important marketing roles, and based on marketing figures of the marketing roles, business capability, performance, business behavior and the like of the marketing roles can be comprehensively evaluated to give suggestions for guiding business operations.

However, in carrying out the concepts of the present disclosure, applicants have discovered that: the prior portrait analysis lacks an effective portrait data intelligent attribution method, and needs to consume a large amount of resources to identify and extract the behavioral characteristics corresponding to the portrait data. In addition, in the process of identifying and extracting the behavior features, only a single behavior feature is considered, and the association between the behavior features and the entities are ignored, so that the key behavior features corresponding to the portrait data cannot be accurately extracted.

BRIEF SUMMARY OF THE PRESENT DISCLOSURE

In view of the above, the present disclosure provides, in one aspect, an image data attribution method, including: acquiring marketing portrait data; calculating an evaluation index score of the marketing portrait according to the marketing portrait data; acquiring behavior data corresponding to the marketing picture with the evaluation index score larger than a preset threshold value; and constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and extracting behavior characteristics influencing the evaluation index score in the behavior data.

According to an embodiment of the present disclosure, the acquiring marketing portrait data includes: and logging in a designated system and/or a Notes mailbox to acquire the marketing portrait data based on a robot process automation technology.

According to an embodiment of the present disclosure, wherein, based on a robot process automation technology, acquiring the marketing portrait data includes: and triggering a data acquisition task at regular time, and distributing the data acquisition task to a client of at least one robot so as to enable the client to download report data and obtain the marketing portrait data.

According to an embodiment of the disclosure, the constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and the extracting behavior features that influence the evaluation index score in the behavior data includes: extracting knowledge from the marketing portrait data and the behavior data to construct a knowledge map of the marketing portrait; analyzing the behavior characteristics in the behavior data by using the knowledge graph, and constructing a behavior characteristic sub-graph and an evaluation index score sub-graph of the marketing portrait; inputting the behavior characteristic subgraph and the evaluation index score subgraph of the marketing portrait into a graph convolution neural network for operation, and outputting the probability distribution of the importance degree of the behavior characteristic to the evaluation index score; and extracting the behavior characteristics of which the probability value is greater than a preset probability threshold.

According to an embodiment of the present disclosure, the inputting the behavior feature sub-graph and the evaluation index score sub-graph of the marketing portrait into the graph convolution neural network for operation includes: and converting the behavior characteristic subgraph and the evaluation index score subgraph of the marketing portrait into a matrix form and inputting the matrix form into the graph convolution neural network for operation.

According to an embodiment of the present disclosure, the attribution method further comprises: acquiring a training data set, wherein the training data set comprises historical marketing portrait data; training the graph convolution neural network using the historic marketing portrait data.

According to an embodiment of the present disclosure, the attribution method further comprises: and guiding marketing operation according to the behavior characteristics of which the probability value is greater than a preset probability threshold.

According to an embodiment of the present disclosure, the attribution method further comprises: after marketing operation is carried out according to the behavior characteristics of which the probability value is greater than a preset probability threshold value, whether the evaluation index score of the marketing portrait is improved or not is judged; if so, increasing the probability value corresponding to the behavior characteristic, and if not, decreasing the probability value corresponding to the behavior characteristic.

According to an embodiment of the present disclosure, the attribution method further comprises: acquiring a training data set, wherein the training data set comprises historical marketing portrait data and marketing portrait data corresponding to marketing portrait with improved evaluation index score; and training the graph convolution neural network by using the historical marketing portrait data and marketing portrait data corresponding to the marketing portrait with the improved evaluation index score.

Another aspect of the present disclosure provides an apparatus for attributing portrait data, comprising: the first acquisition module is used for acquiring marketing portrait data; the calculation module is used for calculating the evaluation index score of the marketing portrait according to the marketing portrait data; the second acquisition module is used for acquiring the behavior data corresponding to the marketing picture with the evaluation index score larger than a preset threshold value; and the extraction module is used for constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and extracting behavior characteristics influencing the evaluation index score in the behavior data.

Another aspect of the present disclosure provides an electronic device including: one or more processors; memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method as described above.

Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions for implementing the method as described above when executed.

Another aspect of the disclosure provides a computer program comprising computer executable instructions for implementing the method as described above when executed.

Drawings

FIG. 1 schematically illustrates a system architecture 100 for an attribution method and apparatus of portrait data according to embodiments of the present disclosure;

FIG. 2 schematically illustrates a flow diagram of an attribution method of portrait data, according to an embodiment of the present disclosure;

FIG. 3 schematically illustrates a flow chart of a marketing representation data acquisition method according to an embodiment of the disclosure;

FIG. 4 schematically shows a flow chart of a behavior feature extraction method according to an embodiment of the present disclosure;

FIG. 5 schematically illustrates a flow diagram of a representation data attribution method according to yet another embodiment of the present disclosure;

FIG. 6 schematically illustrates a flow diagram of a representation data attribution method, according to yet another embodiment of the present disclosure;

FIG. 7 schematically illustrates a flow diagram of a representation data attribution method according to yet another embodiment of the present disclosure;

FIG. 8 schematically illustrates a block diagram of an attribution device of portrait data, according to an embodiment of the present disclosure;

FIG. 9 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure:

FIG. 10 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure;

FIG. 11 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure;

FIG. 12 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure;

FIG. 13 schematically illustrates a block diagram of a first acquisition module according to an embodiment of the present disclosure;

FIG. 14 schematically shows a block diagram of an extraction module according to an embodiment of the present disclosure;

fig. 15 schematically shows a block diagram of an electronic device adapted to implement the above described method according to an embodiment of the present disclosure.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.

Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).

Some block diagrams and/or flow diagrams are shown in the figures. It will be understood that some blocks of the block diagrams and/or flowchart illustrations, or combinations thereof, 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, or other programmable data processing apparatus, such that the instructions, which execute via the processor, create means for implementing the functions/acts specified in the block diagrams and/or flowchart block or blocks. The techniques of this disclosure may be implemented in hardware and/or software (including firmware, microcode, etc.). In addition, the techniques of this disclosure may take the form of a computer program product on a computer-readable storage medium having instructions stored thereon for use by or in connection with an instruction execution system.

An embodiment of the present disclosure provides an attribution method of portrait data, including: marketing portrait data is obtained. And calculating the evaluation index score of the marketing portrait according to the marketing portrait data. And acquiring the behavior data corresponding to the marketing picture with the evaluation index score larger than the preset threshold value. And extracting the behavior characteristics influencing the evaluation index score in the behavior data according to the marketing portrait data and the behavior data.

FIG. 1 schematically illustrates a system architecture 100 for an attribution method and apparatus of portrait data according to embodiments of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.

As shown in fig. 1, the system architecture 100 according to the embodiment may include a storage unit 101, a network 102 and a server 103. Network 102 is used to provide communication links between storage unit 101 and servers 103.

The storage unit 101 may be, for example, a hardware or software implementation, such as an electronic device (e.g., a hard disk) storing data, or a database, made using semiconductor, magnetic media, etc. technologies. The storage unit 101 stores marketing image data, behavior data, and the like of a marketer. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few. The server 103 may be a server capable of acquiring marketing image data of marketers from the storage unit and processing the marketing image data of marketers. According to the embodiment of the present disclosure, in attribution of the image data, the server 103 acquires marketing image data stored on the storage unit 101 through the network 102, and calculates an evaluation index score of the marketing image from the marketing image data. The server 103 acquires behavior data corresponding to the marketing image stored in the storage unit 101 and having the evaluation index score larger than the preset threshold value through the network 102, and extracts behavior features affecting the evaluation index score from the behavior data according to the marketing image data and the behavior data. The server 103 can also give optimization suggestions for guiding marketers to conduct business operations according to the behavior characteristics influencing the evaluation index scores.

It should be noted that the method for attributing portrait data provided by the embodiments of the present disclosure may be executed by the server 103. Accordingly, the image data attribution device provided by the embodiment of the disclosure may be disposed in the server 103. Alternatively, the method of attributing portrait data provided by the embodiments of the present disclosure may also be performed by a server or a server cluster that is different from server 103 and is capable of communicating with storage unit 101 and/or server 103. Accordingly, the image data attributing device provided by the embodiment of the disclosure may also be disposed in a server or a server cluster different from the server 103 and capable of communicating with the storage unit 101 and/or the server 103. Alternatively, the method for attributing portrait data provided by the embodiments of the present disclosure may be performed partly by the server 103 and partly by the storage unit 101. Accordingly, the image data attributing device provided by the embodiment of the present disclosure may also be partially disposed in the server 103 and partially disposed in the storage unit 101.

It should be understood that the number of storage units, networks, and servers in FIG. 1 is illustrative only. There may be any number of storage units, networks, and servers, as desired for an implementation.

The portrait data attribution method provided by the embodiment of the disclosure can be applied to the financial service field, taking a bank as an example, the marketing portrait data generally includes basic information indexes of portrait and performance information index system data, and these data generally include basic information (such as user name, password and the like) of marketing personnel, service specifications (such as customer admission interview completion rate, new contracted customer configuration rate) performance (such as cumulative selection and insurance product configuration amount in the current year, cumulative non-cash private silver product configuration amount in the current year) and the like. The information is deeply mined, the behavior characteristics which play a key role in the indexes can be found, and optimization suggestions can be provided for the marketer based on the behavior characteristics, namely, the current operation behavior of the marketer has places to be improved (for example, the number of outgoing calls is increased). In the process of deep mining of information, the attribution method of the portrait data provided by the embodiment of the disclosure can be adopted to find the behavior characteristics playing key roles in indexes.

It should be understood that the method for attributing portrait data provided by the embodiment of the present disclosure is not limited to be applied to the technical field of financial services, the above description is only exemplary, and for the field related to the attribution of portrait data, such as the sales field of other non-financial products, etc., the method for attributing portrait data according to the embodiment of the present disclosure may be applied to attributing portrait data, so as to accurately extract the key behavior characteristics corresponding to portrait data, and further, may give an optimization suggestion of sales according to the key behavior characteristics, and improve the sales performance of enterprises.

FIG. 2 schematically illustrates a flow diagram of an attribution method of portrait data according to an embodiment of the present disclosure.

As shown in FIG. 2, the attribution method of the image data may include operations S201-S204, for example.

In operation S201, marketing portrait data is acquired.

In operation S202, an evaluation index score of the marketing picture is calculated according to the marketing picture data.

In operation S203, behavior data corresponding to the marketing picture whose evaluation index score is greater than a preset threshold is acquired.

In operation S204, a knowledge graph is constructed and graph convolution is performed on the marketing portrait data and the behavior data, and behavior features affecting the evaluation index score in the behavior data are extracted.

The image data attribution method provided by the embodiment of the disclosure comprises the steps of calculating evaluation index scores of image data, selecting behavior data corresponding to marketing images with the evaluation index scores larger than a preset threshold value to construct a knowledge map and perform graph convolution operation, analyzing the association between behavior characteristics and each behavior characteristic based on the depth of the knowledge map, learning the node characteristics of the knowledge map based on the graph convolution operation, keeping the structural characteristics of the knowledge map as much as possible, avoiding characteristic loss caused by the operation process, and further accurately extracting key behavior characteristics corresponding to the image data. Moreover, the application of the knowledge graph and the graph convolution operation in the extraction of the corresponding behavior characteristics of the portrait data realizes the intelligent attribution of the portrait data.

The method shown in fig. 2 will be further described with reference to the accompanying drawings.

FIG. 3 schematically shows a flow chart of a marketing representation data acquisition method according to an embodiment of the disclosure.

As shown in fig. 3, the marketing image data acquisition method may include operations S301 to S302, for example.

In operation S301, marketing portrait index system data of a marketing user is acquired.

In the embodiment of the present disclosure, the marketing portrait index system data may include, for example, portrait basic information indexes and performance information index system data, including but not limited to, multi-level index names, calculation formulas, calculation weights, and access systems, paths, user names, passwords, etc. of indexes, and if the indexes are obtained through Notes mail information, the indexes also include login mailbox user information and passwords. In a possible situation, the user can customize the index system by himself, that is, the index system can be set by the user through the index customization system. In a specific example of the embodiment of the present disclosure, the obtained index system data may include, for example: the first-level index, the second-level index, a calculation formula of the index, a system address-user name-password, a weight corresponding to the index, and the like may be specifically as shown in table 1 below:

TABLE 1

In operation S302, a designated system and/or Notes mailbox is logged in based on a robot process automation technology, and marketing portrait data is acquired.

In the embodiment of the disclosure, the task of automatically acquiring the data of the index system can be triggered periodically based on a Robot Process Automation (RPA) according to the acquired data of the index system. The method specifically comprises the following steps: based on RPA technique, user information is used to log in the appointed system, and report data is downloaded. Or logging in a Notes mailbox to obtain report data, wherein the report data comprises the required marketing portrait data. The operation of the automatic acquisition task is completed by the coordination of the server, the console and the robot client. The designer realizes RPA execution flow design, and the information comprises: index system, logged-on user, password, acquisition address, NOTES mailbox and report name. After the acquisition process is released, the task scheduling can be triggered periodically through the control end of the user, the task is distributed to the client end of each robot, and the robot completes the execution of the report downloading task according to the established rule. And based on the acquired index bottom layer data, the robot client processes the data again according to an index system, a calculation formula and weight data preset by the console, and calculates the evaluation index score of the marketing portrait.

The marketing portrait data acquisition method provided by the embodiment of the disclosure can realize automatic acquisition of marketing portrait data based on the RPA technology, has high acquisition speed, saves the labor cost consumed by data acquisition, and further realizes intellectualization of portrait data attribution.

Fig. 4 schematically shows a flow chart of a behavior feature extraction method according to an embodiment of the present disclosure.

As shown in fig. 4, the behavior feature extraction method may include, for example, operations S401 to S404.

In operation S401, knowledge extraction is performed on the marketing portrait data and the behavior data to construct a knowledge map of the marketing portrait.

In the embodiment of the present disclosure, in order to ensure the accuracy of the guidance suggestion, behavior data corresponding to the marketing image with the evaluation index score greater than the preset threshold value is obtained, that is, behavior data left in the source system of each underlying index of an excellent marketer when the excellent marketer performs a business operation, for example, the number of calls of the marketer, the number of sound recordings and videos of a product, and the like. After behavior data corresponding to the marketing portrait of the excellent marketing personnel is obtained, the marketing portrait data and the behavior data of the excellent marketing personnel are used as structured and semi-structured data, entities and attributes are extracted through a knowledge extraction technology to be used as nodes of a knowledge graph, and extracted relations are used as edges of the knowledge graph to construct the knowledge graph of the excellent marketing personnel. For example, < salesperson a, marketing task completion rate > may constitute a knowledge-graph triple.

In operation S402, behavior features in the behavior data are analyzed using a knowledge graph, and a behavior feature sub-graph and an evaluation index score sub-graph of the marketing portrait are constructed.

In the embodiment of the disclosure, excellent staff behavior characteristics are deeply analyzed by using the established excellent marketer knowledge graph, and a behavior characteristic sub graph and a score sub graph of each portrait index are established.

In operation S403, the behavior feature sub-graph and the evaluation index score sub-graph of the marketing portrait are input to the graph convolution neural network for computation, and a probability distribution of importance of the behavior feature to the evaluation index score is output.

In the embodiment of the present disclosure, a Graph Convolution Network (GCN) may be used to perform operation processing on the behavior feature sub-graph and the evaluation index score sub-graph of the marketing portrait. The method specifically comprises the following steps: and converting the excellent marketer behavior characteristic subgraph and the portrait index score subgraph into a matrix form, wherein the matrix comprises a characteristic matrix X [ N × D ] formed by behavior characteristics and an adjacent matrix A [ N × N ] of the portrait index score nodes. The matrix X and the matrix A are used as the input of a graph convolution neural network, the graph convolution neural network carries out convolution operation on the matrix X and the matrix A and the like, the probability distribution of the importance degree of the behavior characteristics to the evaluation index score is output, the probability value can be converted into a corresponding characteristic influence factor, and the characteristic influence factor can promote the evaluation index score.

In operation S404, behavior features having probability values greater than a preset probability threshold are extracted.

In the embodiment of the disclosure, by comparing the evaluation index score difference between the marketer to be analyzed and the excellent marketer, the behavior characteristics with the index score influence factor larger than the threshold value are provided as the improvement suggestion. For example, a marketer has a significantly lower client development score at a first level of metrics, a significantly lower retention rate at a second level of compliance clients, and feedback of "increase call out times" and similar attribution and optimization recommendations.

According to the behavior feature extraction method provided by the embodiment of the disclosure, a knowledge graph is constructed on behavior data corresponding to a marketing picture with evaluation index score larger than a preset threshold, the knowledge graph is operated by using a graph convolution neural network, the structural features of the graph are reserved, feature loss caused in the identification process is avoided, the expression vectors of an entity and local neighbor features of the entity can be fully expressed, and further key behavior features influencing evaluation index score of the picture data can be deeply and accurately mined.

FIG. 5 schematically illustrates a flow diagram of a representation data attribution method, according to yet another embodiment of the present disclosure.

As shown in fig. 5, the method may include, for example, operations S501 to S502.

In operation S501, a training data set is obtained, wherein the training data set includes historic marketing representation data.

In operation S502, a convolutional neural network is trained using historical marketing portrait data.

In the embodiment of the disclosure, the marketing portrait data training graph convolutional neural network of the marketing portrait data of the excellent marketers of the historical marketing portrait data can be obtained, so as to ensure the accuracy of behavior feature extraction. The training process may be: marketing portrait data and behavior data of excellent marketers are used as structured and semi-structured data, entities and attributes are extracted through a knowledge extraction technology to serve as nodes of the knowledge graph, and extracted relations serve as edges of the knowledge graph to construct the excellent marketers knowledge graph. And deeply analyzing the behavior characteristics of excellent marketers by using the constructed excellent marketer knowledge graph, constructing behavior characteristic subgraphs and each portrait index score subgraph, and converting the excellent marketer behavior characteristic subgraphs and the portrait index score subgraphs into a matrix form, wherein the matrix form comprises a characteristic matrix formed by behavior characteristics and an adjacent matrix of portrait index score nodes. And taking the matrix as the input of the graph convolution neural network, and training the graph convolution neural network to obtain a trained graph convolution neural network model.

FIG. 6 schematically illustrates a flow diagram of a representation data attribution method, according to yet another embodiment of the present disclosure.

As shown in fig. 6, the method may include, for example, operations S601 to S603.

In operation S601, after the marketing operation is performed according to the behavior feature with the probability value greater than the preset probability threshold, it is determined whether the evaluation index score of the marketing portrait is improved.

If yes, operation S602 is performed, and if no, operation S603 is performed.

In operation S602, a probability value corresponding to the behavior feature is increased.

In operation S603, a probability value corresponding to the behavior feature is reduced.

In the embodiment of the disclosure, increasing the probability value corresponding to the behavior feature may mean that the behavior feature has a particularly obvious effect on guiding the business operation, and the behavior guidance generated based on the behavior feature may greatly improve the sales performance of the marketer. The probability value corresponding to the behavior feature is adjusted to be low, so that the effect of the behavior feature on guiding business operation is general, and the behavior guidance generated based on the behavior feature can enable the sales performance of the marketer not to be greatly improved or not to be improved.

The portrait data attribution method provided by the embodiment of the disclosure can further improve the accuracy of behavior feature extraction by feeding back and adjusting the feature influence factors according to the result after marketing operation.

FIG. 7 schematically illustrates a flow diagram of a representation data attribution method, according to yet another embodiment of the present disclosure.

As shown in fig. 7, the method may include, for example, operations S701 to S702.

In operation S701, a training data set is obtained, wherein the training data set includes historical marketing portrait data and marketing portrait data corresponding to a marketing portrait with an enhanced evaluation index score.

In operation S702, the graph convolutional neural network is trained using historical marketing image data and marketing image data corresponding to the marketing image whose evaluation index score is improved.

In the embodiment of the present disclosure, after the marketer is guided by the behavior characteristics obtained in operation S204, if the marketer evaluation index score is increased and then promoted to be a superior marketer, the superior marketer knowledge map may be included as the latest superior sales. And acquiring the behavior characteristics and the portrait index scores of new excellent marketing personnel and historical excellent marketing personnel as sample data according to the knowledge graph, training the graph convolution neural network model again, and adjusting the influence factors corresponding to the output behavior characteristics. If the image index score is increased after the behavior characteristic related to the improvement suggestion is adopted, which shows that the behavior characteristic has a larger influence on the index score, the influence factor of the behavior characteristic on the index score is increased, otherwise, the influence factor is decreased. And finally, obtaining the optimal influence factor of the behavior characteristics on the image indexes through continuous iterative optimization of the model.

According to the portrait data attribution method provided by the embodiment of the disclosure, the portrait data and the behavior data of new excellent marketing human eyes obtained after marketing operation are combined with the portrait data and the behavior data of historical excellent marketing personnel to carry out iterative training on the atlas neural network, so that the performance of the atlas neural network can be continuously optimized, in the process of carrying out operation on a knowledge atlas based on the atlas neural network with better performance, the feature loss caused in the operation process is further avoided, and the accuracy of feature extraction is further improved.

In summary, the method for attributing portrait data provided by the embodiment of the present disclosure analyzes behavior features and associations between the behavior features based on the depth of the knowledge graph, learns node features of the knowledge graph based on graph convolution operation, and retains structural features of the knowledge graph as much as possible, thereby avoiding feature loss caused by a training process, and further accurately extracting key behavior features corresponding to the portrait data. Moreover, the application of the knowledge graph and the graph convolution operation in the extraction of the corresponding behavior characteristics of the portrait data realizes the intelligent attribution of the portrait data. And moreover, the accuracy of behavior feature extraction can be further improved through continuously and newly generated portrait data and behavior data of excellent marketers to iteratively train the atlas neural network.

FIG. 8 schematically illustrates a block diagram of an attribution device of portrait data, according to an embodiment of the present disclosure.

As shown in FIG. 8, the image data attribution apparatus 800 may comprise, for example, a first obtaining module 810, a calculating module 820, a second obtaining module 830 and an extracting module 840.

The first obtaining module 810 is configured to obtain marketing portrait data.

And the calculating module 820 is used for calculating the evaluation index score of the marketing portrait according to the marketing portrait data.

The second obtaining module 830 is configured to obtain behavior data corresponding to the marketing image with the evaluation index score larger than a preset threshold.

The extraction module 840 is used for constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and extracting behavior characteristics influencing evaluation index scores in the behavior data.

FIG. 9 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure.

As shown in FIG. 9, the representation data attribution device 800 may also include, for example, a training module 850.

A first training module 850 to obtain a training data set, wherein the training data set includes historical marketing representation data. Training the graph convolution neural network using historic marketing profile data.

FIG. 10 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure.

As shown in FIG. 10, the rendering mechanism 800 for representation data may also include, for example, a generation module 860.

And the generating module 860 is used for generating a guidance behavior according to the behavior characteristics that the probability value is greater than the preset probability threshold value, and guiding a service worker to perform service operation according to the guidance behavior.

FIG. 11 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure.

As shown in FIG. 11, the image data attribution apparatus 800 may further include a determining module 870 and an adjusting module 880, for example.

The judging module 870 is configured to judge whether the evaluation index score of the marketing portrait is improved after the marketing operation is performed according to the behavior feature that the probability value is greater than the preset probability threshold.

The adjusting module 880 is configured to increase the probability value corresponding to the behavior feature when the evaluation index score of the marketing portrait is increased, and decrease the probability value corresponding to the behavior feature when the evaluation index score of the marketing portrait is not increased or the increase amplitude is not large.

FIG. 12 schematically illustrates a block diagram of an attribution device of portrait data, according to yet another embodiment of the present disclosure.

As shown in FIG. 12, the representation data attribution device 800 may also include, for example, a second training module 890.

And the second training module 890 is configured to obtain a training data set, where the training data set includes historical marketing portrait data and marketing portrait data corresponding to the marketing portrait with an improved evaluation index score. And training a graph convolution neural network by using the historical marketing image data and the marketing image data corresponding to the marketing image with the improved evaluation index score.

Fig. 13 schematically illustrates a block diagram of a first acquisition module according to an embodiment of the present disclosure.

As shown in fig. 13, the first obtaining module 810 may include, for example, a login unit 811 and a download unit 812.

The login unit 811 is configured to log in a designated system and/or Notes mailbox based on a robot process automation technology.

And the downloading unit 812 is used for triggering the data acquisition task at regular time and distributing the data acquisition task to the client of at least one robot, so that the client downloads report data to obtain marketing portrait data.

FIG. 14 schematically shows a block diagram of an extraction module according to an embodiment of the disclosure.

As shown in fig. 14, the extraction module 840 may include, for example, an extraction unit 841, a construction unit 842, an arithmetic unit 843, and an extraction unit 844.

And the extraction unit 841 is configured to perform knowledge extraction on the marketing portrait data and the behavior data to construct a knowledge graph of the marketing portrait.

The constructing unit 842 is configured to analyze the behavior features in the behavior data by using the knowledge graph, and construct a behavior feature sub-graph and an evaluation index score sub-graph of the marketing portrait.

And an arithmetic unit 843, which is configured to input the behavior feature sub-graph and the evaluation index score sub-graph of the marketing portrait into the graph convolution neural network for arithmetic operation, and output a probability distribution of the importance degree of the behavior feature to the evaluation index score.

An extracting unit 844 is configured to extract the behavior feature with the probability value being greater than a preset probability threshold.

The attribution device for the portrait data, provided by the embodiment of the disclosure, analyzes the behavior characteristics and the association between the behavior characteristics based on the depth of the knowledge graph, learns the node characteristics of the knowledge graph based on graph convolution operation, and retains the structural characteristics of the knowledge graph as much as possible, thereby avoiding the characteristic loss caused by the training process and further accurately extracting the key behavior characteristics corresponding to the portrait data. Moreover, the application of the knowledge graph and the graph convolution operation in the extraction of the corresponding behavior characteristics of the portrait data realizes the intelligent attribution of the portrait data. And moreover, the accuracy of behavior feature extraction can be further improved through continuously and newly generated portrait data and behavior data of excellent marketers to iteratively train the atlas neural network.

Any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware implementations. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.

For example, any number of the first obtaining module 810, the calculating module 820, the second obtaining module 830, the extracting module 840, the first training module 850, the generating module 860, the judging module 870, the adjusting module 880, and the second training module 890 may be combined in one module/unit/sub-unit to be implemented, or any one of the modules/units/sub-units may be split into a plurality of modules/units/sub-units. Alternatively, at least part of the functionality of one or more of these modules/units/sub-units may be combined with at least part of the functionality of other modules/units/sub-units and implemented in one module/unit/sub-unit. According to an embodiment of the present disclosure, at least one of the first obtaining module 810, the calculating module 820, the second obtaining module 830, the extracting module 840, the first training module 850, the generating module 860, the judging module 870, the adjusting module 880, and the second training module 890 may be at least partially implemented as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented by hardware or firmware in any other reasonable manner of integrating or packaging a circuit, or implemented by any one of three implementations of software, hardware, and firmware, or by a suitable combination of any of them. Alternatively, at least one of the first obtaining module 810, the calculating module 820, the second obtaining module 830, the extracting module 840, the first training module 850, the generating module 860, the determining module 870, the adjusting module 880, and the second training module 890 may be at least partially implemented as a computer program module, which when executed, may perform a corresponding function.

It should be noted that the portion of the image data attribution device in the embodiments of the present disclosure corresponds to the portion of the image data attribution method in the embodiments of the present disclosure, and the specific implementation details thereof are the same, and are not repeated herein.

Fig. 15 schematically shows a block diagram of an electronic device adapted to implement the above described method according to an embodiment of the present disclosure. The electronic device shown in fig. 15 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 15, an electronic device 1500 according to an embodiment of the present disclosure includes a processor 1501 which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)1502 or a program loaded from a storage section 1508 into a Random Access Memory (RAM) 1503. Processor 1501 may include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset(s) and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), and so forth. The processor 1501 may also include on-board memory for caching purposes. Processor 1501 may include a single processing unit or multiple processing units for performing different acts of a method flow in accordance with embodiments of the present disclosure.

In the RAM1503, various programs and data necessary for the operation of the electronic apparatus 1500 are stored. The processor 1501, the ROM1502, and the RAM1503 are connected to each other by a bus 1504. The processor 1501 executes various operations of the method flow according to the embodiments of the present disclosure by executing programs in the ROM1502 and/or RAM 1503. Note that the programs may also be stored in one or more memories other than the ROM1502 and RAM 1503. The processor 1501 may also execute various operations of the method flows according to the embodiments of the present disclosure by executing programs stored in the one or more memories.

According to an embodiment of the present disclosure, electronic device 1500 may also include input/output (I/O) interface 1505, input/output (I/O) interface 1505 also being connected to bus 1504. The electronic device 1500 may also include one or more of the following components connected to the I/O interface 1505: an input portion 1506 including a keyboard, a mouse, and the like; an output portion 1507 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 1508 including a hard disk and the like; and a communication section 1509 including a network interface card such as a LAN card, a modem, or the like. The communication section 1509 performs communication processing via a network such as the internet. A drive 1510 is also connected to the I/O interface 1505 as needed. A removable medium 1511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1510 as necessary, so that a computer program read out therefrom is mounted into the storage section 1508 as necessary.

According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 1509, and/or installed from the removable medium 1511. The computer program, when executed by the processor 1501, performs the above-described functions defined in the system of the embodiments of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.

The present disclosure also provides a computer-readable storage medium, which may be contained in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.

According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples may include, but are not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

For example, according to embodiments of the present disclosure, a computer-readable storage medium may include the ROM1502 and/or RAM1503 described above and/or one or more memories other than the ROM1502 and RAM 1503.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.

Claims

1. A method of attributing portrait data, comprising:

acquiring marketing portrait data;

calculating an evaluation index score of the marketing portrait according to the marketing portrait data;

acquiring behavior data corresponding to the marketing picture with the evaluation index score larger than a preset threshold value;

and constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data, and extracting behavior characteristics influencing the evaluation index score in the behavior data.

2. The representation data attribution method of claim 1, wherein the obtaining marketing representation data comprises:

and logging in a designated system and/or a Notes mailbox to acquire the marketing portrait data based on a robot process automation technology.

3. The representation data attribution method of claim 2, wherein the obtaining of the marketing representation data based on robotic process automation technology comprises:

and triggering a data acquisition task at regular time, and distributing the data acquisition task to a client of at least one robot so as to enable the client to download report data and obtain the marketing portrait data.

4. The portrait data attribution method of claim 1, wherein the marketing portrait data and the behavior data are subjected to knowledge graph construction and graph convolution operation, and the extracting of the behavior features affecting the evaluation index score in the behavior data comprises:

extracting knowledge from the marketing portrait data and the behavior data to construct a knowledge map of the marketing portrait;

analyzing the behavior characteristics in the behavior data by using the knowledge graph, and constructing a behavior characteristic sub-graph and an evaluation index score sub-graph of the marketing portrait;

inputting the behavior characteristic subgraph and the evaluation index score subgraph of the marketing portrait into a graph convolution neural network for operation, and outputting the probability distribution of the importance degree of the behavior characteristic to the evaluation index score;

and extracting the behavior characteristics of which the probability value is greater than a preset probability threshold.

5. The method of attributing portrait data as recited in claim 4, wherein the inputting the behavior feature sub-graph and the evaluation index score sub-graph of the marketing portrait into a graph convolution neural network for operation comprises:

and converting the behavior characteristic subgraph and the evaluation index score subgraph of the marketing portrait into a matrix form and inputting the matrix form into the graph convolution neural network for operation.

6. The method of attributing portrait data of claim 4, further comprising:

acquiring a training data set, wherein the training data set comprises historical marketing portrait data;

training the graph convolution neural network using the historic marketing portrait data.

7. The method of attributing portrait data of claim 4, further comprising:

and guiding marketing operation according to the behavior characteristics of which the probability value is greater than a preset probability threshold.

8. The portrait data attribution method of claim 7, further comprising:

after marketing operation is carried out according to the behavior characteristics of which the probability value is greater than a preset probability threshold value, whether the evaluation index score of the marketing portrait is improved or not is judged;

if so, increasing the probability value corresponding to the behavior characteristic, and if not, decreasing the probability value corresponding to the behavior characteristic.

9. The portrait data attribution method of claim 7, further comprising:

acquiring a training data set, wherein the training data set comprises historical marketing portrait data and marketing portrait data corresponding to marketing portrait with improved evaluation index score;

and training the graph convolution neural network by using the historical marketing portrait data and marketing portrait data corresponding to the marketing portrait with the improved evaluation index score.

10. An apparatus for attributing portrait data, comprising:

the first acquisition module is used for acquiring marketing portrait data;

the calculation module is used for calculating the evaluation index score of the marketing portrait according to the marketing portrait data;

the second acquisition module is used for acquiring the behavior data corresponding to the marketing picture with the evaluation index score larger than a preset threshold value;

and the extraction module is used for constructing a knowledge graph and performing graph convolution operation on the marketing portrait data and the behavior data so as to extract behavior characteristics influencing the evaluation index score in the behavior data.

11. An electronic device, comprising:

one or more processors;

a memory for storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-9.

12. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to carry out the method of any one of claims 1 to 9.