CN111915339A - Data processing method, device and equipment - Google Patents

Abstract

Embodiments of the present invention provide a data processing method, device, and device. The method includes: acquiring data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information; The data to be processed is processed to obtain at least one data description information corresponding to the data to be processed, wherein the language model is used for data elements included in the data to be processed and the relationship between each data element Analyze and build a system of rules for the language. By acquiring the data to be processed, and using the language model to process the data to be processed, at least one data description information corresponding to the data to be processed is obtained, which effectively realizes that on the basis of ensuring the cost of data promotion, it is possible to obtain The data describes the information more accurately, and at the same time ensures the promotion or sales effect of the data, further improves the practicability of the method, and is beneficial to the promotion and application of the market.

Description

Data processing method, device and equipment

technical field

The present invention relates to the field of Internet technologies, and in particular, to a data processing method, device and device.

Background technique

With the rapid development of science and technology, the digital technology of information has been gradually advanced in all fields of human life, especially in the fields of e-commerce, Internet finance, logistics, cloud computing and other fields, the application of digital technology of information more extensive.

Taking the e-commerce field as an example, when the data (such as data information, image information, text information) on the e-commerce platform is published, it is necessary to fill in the data description information corresponding to the data, so as to realize the data analysis through the data description information. effective marketing. In the prior art, the data description information is generally edited and designed by professional publishers, or the data description information is filled in by the user. However, when the data description information is edited by professionals, the cost of promotion or promotion is increased; when the user fills in the data description information by himself, the data description information is not accurate enough to attract users, thus reducing the promotion or promotion effect of the data.

SUMMARY OF THE INVENTION

The embodiments of the present invention provide a data processing method, device and device, which can obtain relatively accurate data description information on the basis of ensuring data promotion cost, thereby improving the promotion or promotion effect of data.

In a first aspect, an embodiment of the present invention provides a data processing method, including:

Acquire data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information;

The data to be processed is processed by using a language model to obtain at least one piece of data description information corresponding to the data to be processed, wherein the language model is used for data elements and various data included in the data to be processed The relationships between elements are analyzed and a system of rules for the language is established.

In a second aspect, an embodiment of the present invention provides a data processing device, including:

an acquisition module, configured to acquire data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information;

A processing module, configured to process the data to be processed by using a language model to obtain at least one data description information corresponding to the data to be processed, wherein the language model is used to process the data included in the data to be processed Data elements and the relationship between each data element are analyzed, and a language rule system is established.

In a third aspect, an embodiment of the present invention provides an electronic device, including: a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are processed by the The data processing method in the first aspect above is implemented when the server is executed.

In a fourth aspect, an embodiment of the present invention provides a computer storage medium for storing a computer program, and the computer program enables a computer to implement a data processing method in the first aspect above.

By acquiring the data to be processed, and using the language model to process the data to be processed, at least one data description information corresponding to the data to be processed is obtained, which effectively realizes that on the basis of ensuring the cost of data promotion, it is possible to obtain The data describes the information more accurately, and at the same time ensures the promotion or sales effect of the data, further improves the practicability of the method, and is beneficial to the promotion and application of the market.

In a fifth aspect, an embodiment of the present invention provides a data processing method, including:

determine the language model;

Input the object to be processed and the data to be processed corresponding to the object to be processed into the language model to obtain at least one promotion description information corresponding to the object to be processed, and the data to be processed includes at least one of the following : text information, picture information, video information;

Wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between the various data elements to obtain the element relationship Result: The rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship result to determine the output data of the language model.

In a sixth aspect, an embodiment of the present invention provides a data processing device, including:

A determination module for determining a language model;

The input module is used to input the object to be processed and the data to be processed corresponding to the object to be processed into the language model, and obtain at least one promotion description information corresponding to the object to be processed, the data to be processed Including at least one of the following: text information, picture information, video information;

Wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between the various data elements to obtain the element relationship Result: The rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship result to determine the output data of the language model.

In a seventh aspect, an embodiment of the present invention provides an electronic device, including: a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are processed by the The data processing method according to the fifth aspect above is implemented when the processor is executed.

In an eighth aspect, an embodiment of the present invention provides a computer storage medium for storing a computer program, and the computer program enables the computer to implement the data processing method described in the fifth aspect above.

By determining the language model, input the object to be processed and the data to be processed corresponding to the object to be processed into the language model, and then use the language model to process the data information to obtain at least one corresponding to the object to be processed. A publicity description information, which can be used to promote the object to be processed, which effectively realizes that on the basis of ensuring the cost of data sales, more accurate publicity and description information can be obtained, ensuring the promotion or sales of the object to be processed. effect, further improve the practicability of the method, and is beneficial to the promotion and application of the market.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

1a is a schematic diagram of a system structure of a method for processing application data according to an embodiment of the present invention;

FIG. 1 is a flowchart 1 of a data processing method provided by an embodiment of the present invention;

2 is a schematic interface diagram of a data description information provided by an embodiment of the present invention;

3 is a second flowchart of a data processing method provided by an embodiment of the present invention;

4 is a third flowchart of a data processing method provided by an embodiment of the present invention;

5 is a fourth flowchart of a data processing method provided by an embodiment of the present invention;

6 is a flowchart of another data processing method provided by an embodiment of the present invention;

7 is a flowchart 1 of a data processing method provided by an application embodiment of the present invention;

FIG. 8 is a second flowchart of a data processing method provided by an application embodiment of the present invention;

9 is a third flowchart of a data processing method provided by an application embodiment of the present invention;

10 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present invention;

11 is a schematic structural diagram of an electronic device corresponding to the data processing apparatus provided by the embodiment shown in FIG. 10;

12 is a schematic structural diagram of another data processing apparatus provided by an embodiment of the present invention;

FIG. 13 is a schematic structural diagram of an electronic device corresponding to the data processing apparatus provided by the embodiment shown in FIG. 12 .

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a," "the," and "the" as used in the embodiments of the present invention and the appended claims are intended to include the plural forms as well, unless the context clearly dictates otherwise, "a plurality" Generally, at least two kinds are included, but the case of including at least one kind is not excluded.

It should be understood that the term "and/or" used in this document is only an association relationship to describe the associated objects, indicating that there may be three kinds of relationships, for example, A and/or B, which may indicate that A exists alone, and A and B exist at the same time. B, there are three cases of B alone. In addition, the character "/" in this document generally indicates that the related objects are an "or" relationship.

Depending on the context, the words "if", "if" as used herein may be interpreted as "at" or "when" or "in response to determining" or "in response to detecting". Similarly, the phrases "if determined" or "if detected (the stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detected (the stated condition or event)," depending on the context )" or "in response to detection (a stated condition or event)".

It should also be noted that the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion, such that a commodity or system comprising a list of elements includes not only those elements, but also includes not explicitly listed other elements, or elements inherent to the commodity or system. Without further limitation, an element defined by the phrase "comprising a..." does not preclude the presence of additional identical elements in the article or system that includes the element.

In addition, the sequence of steps in the following method embodiments is only an example, and is not strictly limited.

Definition of Terms:

The language model is an abstract mathematical modeling of language based on the objective facts of the language, and it is a kind of correspondence; the relationship between the language model and the objective facts of the language is like the relationship between the abstract straight line and the concrete straight line in mathematics. There are three main types of language models: generative models, analytical models and identification models; specifically, generative models start from a formal language system and generate a certain set of languages, such as N. Chomsky's formal language Theory and transformation grammar. Analytical model starts from a certain set of language, clarifies the relationship between these elements according to the analysis of the properties of each element in this set, and on this basis uses deductive methods to establish a language rule system, such as Soviet mathematics. The language model proposed by O.C. Kuragina and Romanian mathematician S. Marcus using the method of set theory. On the basis of the generative model and the analytical model, combining the two produces a model with great practical value, that is, the identification model. The identification model can start from a certain set of language elements and a rule system, and determine whether these elements are certain words or qualified sentences in the language through the operation of limited steps. Such as Y. Barr-Hiller proposed a syntactic type calculus model by mathematical logic method.

In order to facilitate the understanding of the technical solutions of the present application, the prior art is briefly described by taking the field of e-commerce as an example:

In the prior art, the ability to automatically generate relevant product information descriptions based on a little information (pictures and text, etc.) provided by users has gradually attracted the attention of the industry. For example, Boltzmann Machine (VAE, Boltzmann Machine) generation technology can be used. The data is processed to obtain the description of commodity information; then, the Boltzmann machine generation technology is replaced by the adversarial generation technology (Generative Adversarial Networks, referred to as GAN). However, when using GAN technology to analyze data, there are the following problems:

1) GAN technology has achieved good results in image style transfer, such as Pix2Pix GAN and Cycle GAN for Image-to-image, but GAN technology is not yet mature in processing video data;

2) The effect of GAN technology alone on text generation is limited, but after combining with reinforcement learning algorithms, such as: Seq GAN has achieved good results. Therefore, how to train GAN to generate accurate, rich, diverse and personalized product descriptions is a great challenge;

3) Automatically generate a series of related product description information (poster, main image, title, description) by integrating multi-modal information (picture, video, text), at the same time consider the buyer's attributes for personalized generation, and consider the season to make relevant recommendations is a difficult problem.

In order to solve the above technical problems, referring to FIG. 1a, the present embodiment provides a data processing system that can implement a data processing method, the data processing system includes a user terminal 01 and a processing device 02 that is communicatively connected to the user terminal 01, Wherein, the user terminal 01 may be various electronic devices such as a mobile phone, a tablet computer, a personal computer, etc., and the processing device 02 may be implemented as software, or a combination of software and hardware. In addition, the user terminal 01 and the processing device 02 may be a client/server CS architecture built on a local area network; alternatively, the user terminal 01 and the processing device 02 may also be a browser/server BS architecture built on a wide area network, skilled in the art Personnel can choose different implementation methods according to specific application scenarios. specific:

User terminal 01, configured to detect the execution operation input by the user, and generate a pending request based on the execution operation, the pending request may include pending data, wherein the pending data may include at least one of the following: text information, picture information , video information. After the user terminal 01 obtains the pending request input by the user, the pending request may be sent to the processing device 02 .

The user terminal 01 may be provided with an application program or application website corresponding to the processing device 02, and the user can send the input pending request to the processing device 02 through the application program or the application website; for example, in the user When data prediction processing is desired, a pending request can be input through the data input port in the application program. After the user terminal 01 obtains the aforementioned pending request, the pending request can be sent to the processing device 02 for processing.

The processing device 02 is connected in communication with the user terminal 01, and is used for receiving the pending request sent by the user terminal 01. After receiving the pending request, the language model can be used to process the pending data to obtain at least one corresponding to the pending data. A piece of data description information, the data description information at this time includes at least one of the following: image information, title, data description text, and data description video.

In this embodiment, after the pending request is obtained, the corresponding pending data can be obtained through the pending request, and then the language model is used to process the pending data to obtain at least one data description information corresponding to the pending data, thereby obtaining at least one data description information corresponding to the pending data. The data description information can be automatically generated according to the information included in the data to be processed, which effectively realizes the acquisition of more accurate product description information on the basis of ensuring the cost of product promotion, thereby improving the promotion or sales effect of the product and satisfying the Therefore, the practicability of the method is improved, which is beneficial to the promotion and application of the market.

In order to facilitate the understanding of the technical solutions in this embodiment, the data processing process is described in detail below. Referring to FIG. 1 , this embodiment provides a data processing method, and the execution body of the processing method is a data processing method. Processing means, which can be implemented as software, or a combination of software and hardware. Specifically, the method may include:

S101: Acquire data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information.

The video information may include picture information and text information. However, this embodiment does not limit the specific implementation of acquiring the data to be processed, and those skilled in the art can set it according to specific usage requirements. For example, the data to be processed can be stored in a preset area, and can be acquired by accessing the preset area. to the data to be processed. Alternatively, the data to be processed can also be stored in other devices (for example, a user terminal), and the data to be processed stored in the device can be obtained through a communication connection with the device. Specifically, an achievable method is: : A data acquisition request can be sent to the device, so that the device can return the data to be processed based on the data request; or, another achievable way is: the device can actively send the data to be processed to the processing device, so that the processing device can Data to be processed can be obtained actively.

It can be understood that, when the device end is a user terminal, after the user terminal obtains the data to be processed, preferably, the user terminal can actively send the data to be processed to the processing device for processing, thereby improving the quality of data processing. and efficiency.

S102: Use a language model to process the data to be processed, and obtain at least one data description information corresponding to the data to be processed, wherein the language model is used for data elements included in the data to be processed and between data elements The relationship is analyzed and the rule system of the language is established.

The language model may be an analytical language model or a discriminative language model, and the language model is used to process the data to be processed. Analyze the relationship and establish a language rule system. In the computer field, the smallest unit of information is a bit of a binary number, also called a bit; and a bit string formed by combining several bits represents a data Elements (such as: representing an integer with a word-length bit string, representing a character with an 8-bit binary number, etc.), this bit string is usually called an element or a node. Therefore, after obtaining the data elements included in the data to be processed, the language model can be used to analyze the properties of the data elements included in the data to be processed, determine the relationship between the various data elements, and based on the above information, Use the deductive method to establish the rule system of the language.

It should be noted that the language model can directly and effectively process the data to be processed including video information, so that at least one data description information corresponding to the data to be processed can be obtained. The data description information can include at least the following: One: image information, title, data description text, data description video; through the generated data description information, the promotion or sales effect of the data can be effectively improved.

For example, a set of data to be processed about women's clothing is now obtained, and the data to be processed at this time includes text information and picture information for the women's clothing. After the above-mentioned data to be processed is obtained, the language model can be used to process the data. By performing analysis and processing, at least one data description information corresponding to the above-mentioned women's clothing can be obtained, which effectively realizes that more accurate data description information can be obtained on the basis of ensuring the cost of data promotion, and then the generated data description information can be used. Improve the promotion or sales effect of data.

Specifically, as shown in FIG. 2 , which is a schematic diagram of a display interface of data description information, the data description information in the display interface may include: title "Where is it", title description information, text title "style", text title The description information and the image information (for example, posters, pictures, etc.) corresponding to the picture information in the data to be processed, or the data describes the video, and so on. It can be understood that the style or layout of the display interface of the data description information is not limited to the style or layout of the display interface shown in the above Figure 2, and those skilled in the art can also arbitrarily set according to specific design requirements and application requirements. This will not be repeated here.

Optionally, after obtaining at least one data description information corresponding to the data to be processed, the method in this embodiment further includes:

S103: Output at least one data description information.

After obtaining the data description information, the data description information can be selectively output according to the user's needs. For example, when the user requests to output the data description information, at least one data description information can be output according to the user's request, so that the user can intuitively and quickly The content of the data description information can be obtained by the user; when the user does not request to output the data description information, the operation of outputting the data description information may not be performed, which effectively improves the flexibility of the method.

In the data processing method provided in this embodiment, by acquiring the data to be processed, and processing the data to be processed by using a language model, at least one data description information corresponding to the data to be processed is obtained, which effectively realizes the cost reduction of data promotion. On the basis, more accurate data description information can be obtained, meanwhile, the promotion or sales effect of the data can be ensured, and the practicability of the method can be further improved, which is beneficial to the promotion and application of the market.

FIG. 3 is a second flow chart of a data processing method provided by an embodiment of the present invention; on the basis of the above-mentioned embodiment, and continuing to refer to FIG. 3 , before acquiring the data to be processed, the method in this embodiment also further Can include:

S001: Acquire training data and training description information corresponding to the training data.

The training data may include at least one of the following: processed data for historical objects, feedback data for historical objects, and model output data generated for historical objects. Correspondingly, the training description information may include at least one of the following: description information corresponding to the processed data, description information fed back for the historical object, and description information generated for the historical object. It can be understood that the historical objects may refer to: commodities, services or other similar objects and so on.

In addition, this embodiment does not limit the specific implementation of acquiring training data and training description information corresponding to the training data, and those skilled in the art can set according to specific usage requirements, for example: training data and training data corresponding to the training data. The training description information can be stored in the preset area, and the training data and the training description information corresponding to the training data can be obtained by accessing the preset area. Alternatively, the training data and the training description information corresponding to the training data may also be stored in other devices, and the training data stored in the device and the training description corresponding to the training data are acquired through a communication connection with the device. information.

Of course, those skilled in the art can also obtain the training data and the training description information corresponding to the training data in other ways, as long as the accuracy and reliability of the acquisition of the training data and the training description information can be ensured, which will not be repeated here. .

S002: Learning and training the training data and the training description information to obtain a language model.

After the training data and the training description information are obtained, the training data and the training description information can be learned and trained, thereby obtaining a language model, which can analyze and identify each element of the training data and the training description information, so as to obtain relationship between elements.

Obtaining the language model in the above manner effectively ensures the accuracy and reliability of the language model acquisition, and further improves the processing quality and efficiency of the method.

Optionally, based on the implementation manner of the foregoing embodiment, after obtaining at least one data description information corresponding to the data to be processed, the method in this embodiment may further include:

S201: Adjust the language model by using training data, training description information corresponding to the training data, data to be processed, and at least one data description information.

For example, in the existing data A to be processed, after using the training data and the training description information to obtain a language model, the language model can be used to analyze and process the data A to be processed, so that the data description information A corresponding to the data A to be processed can be obtained. , after obtaining the data description information A, the training data, training description data, pending data A and data description information A can be used again to adjust and update the language model, thereby improving the accuracy and reliability of the data processing by the language model, and further Improves the quality and efficiency of language model processing of other data.

Fig. 4 is a third flowchart of a data processing method provided by an embodiment of the present invention; on the basis of any of the above-mentioned embodiments, and continuing to refer to Fig. 4, after obtaining at least one data corresponding to the data to be processed After describing the information, the method in this embodiment may further include:

S301: Acquire behavior characteristics of the user.

Wherein, the behavior characteristics of the user may include at least one of the following: the behavior of the user operating the network, the behavior of the user logging in to the network, and the behavior of the user accessing the network. In addition, this embodiment does not limit the specific acquisition method of the user's behavior characteristics, and those skilled in the art can arbitrarily set according to specific design requirements and application requirements. Obtain user behavior characteristics and so on.

S302: Determine target description information from at least one piece of data description information according to the behavioral characteristics of the user.

After acquiring the user's behavioral characteristics, target description information can be determined in at least one data description information according to the user's behavioral characteristics. For example, the user's preferences, interests and other information can be analyzed based on the user's behavioral characteristics The user's preferences or interests can be used to identify the matching degree of at least one data description information, and the data description information whose matching degree is greater than or equal to a preset threshold can be determined as the target description information. It can be understood that the target description information can be one or more. In this way, data description information (ie, target description information) that is more in line with user needs can be obtained, which further improves the accuracy of the method.

Optionally, the method in this embodiment may further include:

S303: Output target description information.

After obtaining the target description information, the target description information can be selectively output according to the user's needs. For example, when the user requests to output the target description information, at least one target description information can be output according to the user's request, so that the user can intuitively and quickly The content of the target description information can be obtained in a timely manner; when the user does not request to output the target description information, the operation of outputting the target description information may not be performed, which effectively improves the flexibility of the method.

FIG. 5 is a fourth flowchart of a data processing method provided by an embodiment of the present invention; on the basis of any of the above-mentioned embodiments, with continued reference to FIG. 5 , the method in this embodiment may further include:

S401: Acquire feedback information sent by a user for the data description information.

S402: Adjust the language model according to the feedback information.

The data description information in this embodiment may be at least one piece of data description information obtained directly after processing the data to be processed by using a language model, or the data description information may also refer to at least one piece of data description information after analysis and processing. The obtained target description information. That is, an achievable way is: after obtaining at least one data description information corresponding to the data to be processed, the user can send feedback information for the data description information through the terminal device, and the feedback information can include accurate data description information. The language model can be adjusted according to the feedback information after the feedback information sent by the user for the data description information is obtained. Another achievable way is: after determining the target description information in at least one data description information according to the user's behavioral characteristics, the user can send feedback information for the target description information through the terminal device, and the feedback information can include the target description information. Accuracy description or standard data description information for the data to be processed, etc. After obtaining the feedback information sent by the user for the target description information, the language model can be adjusted according to the feedback information.

Specifically, when adjusting the language model according to the feedback information, an achievable way is: after obtaining the language model based on the training data and the training description information, the language model can be adjusted according to the feedback information; The language model is adjusted by the data to be processed, at least one piece of data description information corresponding to the data to be processed, and feedback information. Still alternatively, another achievable manner is: after the language model is obtained based on the training data and the training description information, the language model can be adjusted according to the training data, the training description information, and the feedback information. Or, another achievable way is: after obtaining the language model based on the training data and the training description information, the training data, the training description information, the feedback information, the data to be processed, and at least one piece of data corresponding to the data to be processed can be obtained. The description information adjusts the language model.

Those skilled in the art can choose different ways to adjust and update the language model according to specific application requirements, as long as the language model can be adjusted and updated in time, which can not only improve the accuracy and reliability of data processing by the language model , which can further improve the quality and efficiency of the language model processing other data.

FIG. 6 is a flowchart of another data processing method provided by an embodiment of the present invention; with reference to FIG. 6 , this embodiment provides another data processing method, and the execution body of the processing method is data processing means, the processing means may be implemented as software, or a combination of software and hardware. Specifically, the method may include:

S501: Determine the language model.

The language model may include a regression estimation sub-algorithm and a rule inference sub-algorithm, wherein the regression estimation sub-algorithm is a statistical analysis method for determining the interdependent quantitative relationship between two or more variables; in this embodiment, the The regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between each data element to obtain the element relationship result. In addition, the rule inference sub-algorithm refers to the formal description of the expert knowledge in the relevant field to form system rules. In this embodiment, the rule inference sub-algorithm is used to establish a language rule system, and the rule system is used for all The element relationship results are analyzed and processed to determine the output data of the language model.

S502: Input the object to be processed and the data to be processed corresponding to the object to be processed into the language model to obtain at least one promotion description information corresponding to the object to be processed, and the data to be processed includes the following at least One: text information, picture information, video information.

The objects to be processed may include commodities and/or services, for example, the objects to be processed may be clothing, products, etc. to be sold and promoted, or the objects to be processed may also be catering services, entertainment services, etc. to be sold. The data information is processed by using a language model to obtain at least one promotion description information corresponding to the object to be processed, wherein the language model is used to analyze the data elements included in the data to be processed and each data element The relationship between them is analyzed and a system of language rules is established.

The implementation process and implementation effect of the above steps in this embodiment are similar to the implementation process and implementation effect of steps S101-S102 in the above embodiment. For details, reference may be made to the above statement, which will not be repeated here.

It is conceivable that the method in this embodiment may also include some or all of the methods in the above-mentioned embodiments shown in FIG. 2 to FIG. 6 . related descriptions of the examples. For the execution process and technical effects of the technical solution, refer to the descriptions in the embodiments shown in FIG. 2 to FIG. 6 , which will not be repeated here.

In the data processing method provided in this embodiment, by determining a language model, input the object to be processed and the data to be processed corresponding to the object to be processed into the language model, and then use the language model to process the data information to obtain At least one piece of publicity description information corresponding to the object to be processed, the publicity description information can be used to promote the object to be processed, effectively realizing that on the basis of ensuring the cost of data promotion, more accurate publicity description information can be obtained. , which ensures the promotion or sales effect of the object to be treated, further improves the practicability of the method, and is beneficial to the promotion and application of the market.

In a specific application, referring to FIGS. 7-9 , this application embodiment provides a data processing method. The data processing method includes training to obtain a language model and a process of using the language model. For details, refer to the accompanying drawings. As shown in Figure 7, the steps of training to obtain a language model include:

step1: get training data

The training data includes at least one of the following: training data a - product and description data pair edited by the user, training data b - product and description data pair produced according to online user feedback, and training data c - current model output Commodities and description data pairs. Specifically, the training data a is a product description and description data pair edited for historical products. After the training data a is obtained, the language model can be obtained after the training data a can be learned and trained separately, and the training data b is the user Feedback information provided for the data description information output by the language model, the feedback information may include data information and data description information for the product to be processed.

Step 2: Perform model training based on the training data to obtain a language model.

Specifically, in the process of model training, it may include using a preset algorithm (for example, a reward estimation model) to perform multi-dimensional estimation processing on the training data, and the estimation processing is used to ensure the diversity and richness of the training data. description, and can also analyze and process the correlation of the training data, so as to obtain the correlation score between the training data. Further, the GAN discriminator can also be used to evaluate the training data, so as to improve the identification of the training data. accuracy.

It is understandable that, after the language model is obtained, the error back propagation algorithm can also be used to adjust and update the model parameters in the language model in time, so as to further improve the accuracy of the language model.

With reference to accompanying drawing 8-9, the step that utilizes language model to process data comprises:

Step 101 : external data collection, wherein the external data refers to the training data in the above drawings, the external data collection is used to implement the training process of the language model, and the language model can process the data to be processed, thereby outputting commodity description information.

Step 102: Perform language model training based on the collected external data.

Input: integrated training data;

Output: language model.

Step 103: Use the language model to process the data to be processed, so as to generate data description information (for example: commodity description information).

Input: data information (for example: product information), language model;

Output: at least one data description information corresponding to the data.

step104: Personalized distribution of data description.

Input: user information, data description;

Output: Recommended user data description information.

step105: User online data recovery.

Output: The user's feedback on the presentation description, which can be used to optimize and update the language model.

Wherein, referring to Figure 9, the step 103 may specifically include the following processes:

Step 1031: Acquire commodity information to be processed, the commodity information may include: commodity image, commodity copy (text) or commodity video and so on.

Step 1032: Input the commodity information to be processed into the language model, so as to realize the analysis and identification of the commodity information, and output a commodity description set corresponding to the commodity information.

Step 1033: Use the preset correlation model to analyze and process the commodity description set, so as to obtain a high-precision commodity description set; specifically, the correlation model can correlate the commodity description information in the commodity description set with the user's behavior. Sorting can obtain commodity description information that is highly relevant to the user's behavior, and further improves the recommendation for the user. More accurate commodity description information.

Step 1034: Output the commodity description set, output the final high-precision commodity description set, so as to facilitate the user to view it directly.

In the data processing method provided by this application embodiment, the data to be processed is obtained by acquiring the data to be processed, wherein the data to be processed may include text information, picture information, and video information, and the language model is used to process the data to be processed, and the data corresponding to the data to be processed is obtained. At least one piece of data description information, thus overcoming the immature defect of GAN technology in video processing in the prior art; in addition, the product description set is analyzed and processed through the correlation model, so that a high-precision product description set can be obtained, effectively It realizes the recommendation of different commodity description information for different users, and realizes the personalized recommendation information for users. In addition, this method also effectively realizes that more accurate commodity description information can be obtained on the basis of guaranteeing the cost of commodity promotion. , which ensures the promotion or sales effect of the product, further improves the practicability of the method, and is beneficial to the promotion and application of the market.

FIG. 10 is a schematic structural diagram of a data processing apparatus provided by an embodiment of the present invention; with reference to FIG. 10 , the present embodiment provides a data processing apparatus, and the processing apparatus can execute the processing corresponding to FIG. 1 above. data processing method. Specifically, the processing device may include:

The obtaining module 11 is configured to obtain data to be processed, and the data to be processed includes at least one of the following: text information, picture information, and video information;

The processing module 12 is configured to process the data to be processed by using a language model to obtain at least one data description information corresponding to the data to be processed, wherein the language model is used to analyze the data elements included in the data to be processed and each The relationships between data elements are analyzed and a system of rules for the language is established.

Optionally, the data description information includes at least one of the following: image information, title, and data description text.

Optionally, before acquiring the data to be processed, the acquiring module 11 and the processing module 12 in this embodiment can also be used to perform the following steps:

an acquisition module 11, for acquiring training data and training description information corresponding to the training data;

The processing module 12 is used for learning and training the training data and the training description information to obtain a language model.

Optionally, after obtaining at least one data description information corresponding to the data to be processed, the method further includes:

The language model is adjusted by using training data, training description information corresponding to the training data, data to be processed, and at least one data description information.

Optionally, after obtaining at least one piece of data description information corresponding to the data to be processed, the processing module 12 in this embodiment may be further configured to perform the following step: output at least one piece of data description information.

Optionally, after obtaining at least one piece of data description information corresponding to the data to be processed, the obtaining module 11 and the processing module 12 in this embodiment may also be configured to perform the following steps:

an acquisition module 11, used to acquire the behavioral characteristics of the user;

The processing module 12 is configured to determine target description information in at least one piece of data description information according to the behavioral characteristics of the user.

Optionally, the processing module 12 in this embodiment may also be configured to perform the following steps: outputting target description information.

Optionally, the acquiring module 11 and the processing module 12 in this embodiment can also be used to perform the following steps:

The obtaining module 11 is used to obtain the feedback information sent by the user for the data description information;

The processing module 12 is configured to adjust the language model according to the feedback information.

The apparatus shown in FIG. 10 can execute the methods of the embodiments shown in FIGS. 1-5 and 7-9. For the parts not described in detail in this embodiment, please refer to the methods shown in FIGS. 1-5 and 7-9. Relevant description of the embodiment. For the execution process and technical effects of the technical solution, refer to the descriptions in the embodiments shown in FIGS. 1-5 and 7-9 , which will not be repeated here.

In a possible design, the structure of the data processing apparatus shown in FIG. 10 may be implemented as an electronic device, and the electronic device may be various devices such as a mobile phone, a tablet computer, and a server. As shown in FIG. 11 , the electronic device may include: a first processor 21 and a first memory 22 . Wherein, the first memory 22 is used to store programs that support the electronic device to execute the data processing methods provided in the embodiments shown in FIGS. 1-8 , and the first processor 21 is configured to execute the programs stored in the first memory 22 program of.

The program includes one or more computer instructions, wherein, when the one or more computer instructions are executed by the first processor 21, the following steps can be implemented:

Acquire data to be processed, and the data to be processed includes at least one of the following: text information, picture information, and video information;

The data to be processed is processed by using the language model, and at least one piece of data description information corresponding to the data to be processed is obtained.

Optionally, the first processor 21 is further configured to execute all or part of the steps in the foregoing embodiments shown in FIGS. 1-5 and 7-9 .

The structure of the electronic device may further include a first communication interface 23 for the electronic device to communicate with other devices or a communication network.

In addition, an embodiment of the present invention provides a computer storage medium for storing computer software instructions used by an electronic device, which includes data for executing the data in the method embodiments shown in FIGS. 1-5 and 7-9. The procedures involved in the processing method.

FIG. 12 is a schematic structural diagram of another data processing apparatus provided by an embodiment of the present invention; with reference to FIG. 12 , this embodiment provides another data processing apparatus, and the processing apparatus can execute the above-mentioned FIG. 6 The corresponding data processing method. Specifically, the processing device may include:

The determination module 31 is used to determine a language model; wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used for data elements included in the data to be processed and each data element. The relationship between the elements is analyzed to obtain element relationship results; the rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship results to determine the output of the language model. data.

The input module 32 is used to input the object to be processed and the data to be processed corresponding to the object to be processed into the language model, and obtain at least one promotion description information corresponding to the object to be processed. The data includes at least one of the following: text information, picture information, and video information.

The apparatus shown in FIG. 12 can execute the method of the embodiment shown in FIG. 6 to FIG. 9 . For parts not described in detail in this embodiment, reference may be made to the related description of the embodiment shown in FIG. 6 to FIG. 9 . For the execution process and technical effects of the technical solution, refer to the descriptions in the embodiments shown in FIG. 6 to FIG. 9 , which will not be repeated here.

In a possible design, the structure of the data processing apparatus shown in FIG. 12 may be implemented as an electronic device, and the electronic device may be various devices such as a mobile phone, a tablet computer, and a server. As shown in FIG. 13 , the electronic device may include: a second processor 41 and a second memory 42 . Wherein, the second memory 42 is used to store a program that supports the electronic device to execute the data processing method provided in the embodiment shown in FIG. 6 , and the second processor 41 is configured to execute the program stored in the second memory 42 .

The program includes one or more computer instructions, wherein the one or more computer instructions can realize the following steps when executed by the second processor 41:

determine the language model;

Input the object to be processed and the data to be processed corresponding to the object to be processed into the language model to obtain at least one promotion description information corresponding to the object to be processed, and the data to be processed includes at least one of the following : text information, picture information, video information;

Wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between the various data elements to obtain the element relationship Result: The rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship result to determine the output data of the language model.

Optionally, the second processor 41 is further configured to execute all or part of the steps in the foregoing embodiments shown in FIG. 7 to FIG. 9 .

The structure of the electronic device may further include a second communication interface 43 for the electronic device to communicate with other devices or a communication network.

In addition, an embodiment of the present invention provides a computer storage medium for storing computer software instructions used by an electronic device, which includes a program for executing the data processing method in the method embodiment shown in FIG. 6-FIG. 9. .

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by adding a necessary general hardware platform, and certainly can also be implemented by combining hardware and software. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of computer products in essence or that contribute to the prior art. In the form of a computer program product embodied on a medium (including but not limited to disk storage, CD-ROM, optical storage, etc.).

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block in 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 device to produce a machine such that the instructions executed by the processor of the computer or other programmable device produce a process for implementing the process. Figures a process or processes and/or block diagrams of a block or devices of the functions specified in the blocks.

The computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable device to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means that implements A function specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be loaded onto a computer or other programmable device such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide Steps for implementing the function specified in one or more of the flowcharts and/or one or more blocks of the block diagrams.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

Memory may include non-persistent memory in computer readable media, random access memory (RAM) and/or non-volatile memory in the form of, for example, read only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media includes both persistent and non-permanent, removable and non-removable media, and storage of information may be implemented by any method or technology. Information may be computer readable instructions, data structures, modules of programs, or other data. Examples of computer storage media 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 Disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, excludes transitory computer-readable media, such as modulated data signals and carrier waves.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (13)

1. a processing method of data, is characterized in that, comprises: Acquire data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information; The data to be processed is processed by using a language model to obtain at least one piece of data description information corresponding to the data to be processed, wherein the language model is used for data elements and various data included in the data to be processed The relationships between elements are analyzed and a system of rules for the language is established. 2. The method according to claim 1, wherein the data description information comprises at least one of the following: Image information, title, data description text. 3. The method according to claim 1, wherein before acquiring the data to be processed, the method further comprises: obtaining training data and training description information corresponding to the training data; Learning and training the training data and training description information to obtain the language model. 4. The method according to claim 3, wherein after obtaining at least one data description information corresponding to the data to be processed, the method further comprises: The language model is adjusted by using the training data, training description information corresponding to the training data, the data to be processed, and at least one piece of the data description information. 5. The method according to any one of claims 1-4, wherein after obtaining at least one data description information corresponding to the data to be processed, the method further comprises: At least one of the data description information is output. 6. The method according to any one of claims 1-4, wherein after obtaining at least one data description information corresponding to the data to be processed, the method further comprises: Obtain the behavioral characteristics of users; Target description information is determined in at least one of the data description information according to the behavior characteristics of the user. 7. The method according to claim 6, wherein the method further comprises: Output the target description information. 8. The method according to any one of claims 1-4, wherein the method further comprises: Obtain the feedback information sent by the user for the data description information; The language model is adjusted according to the feedback information. 9. A data processing device, comprising: an acquisition module, configured to acquire data to be processed, where the data to be processed includes at least one of the following: text information, picture information, and video information; A processing module, configured to process the data to be processed by using a language model to obtain at least one data description information corresponding to the data to be processed, wherein the language model is used to process the data included in the data to be processed Data elements and the relationship between each data element are analyzed, and a language rule system is established. 10. An electronic device, comprising: a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein when the one or more computer instructions are executed by the processor A data processing method as claimed in any one of claims 1 to 8 is implemented. 11. A method for processing data, comprising: determine the language model; Input the object to be processed and the data to be processed corresponding to the object to be processed into the language model to obtain at least one promotion description information corresponding to the object to be processed, and the data to be processed includes at least one of the following : text information, picture information, video information; Wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between the various data elements to obtain the element relationship Result: The rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship result to determine the output data of the language model. 12. A data processing device, comprising: A determination module for determining a language model; The input module is used to input the object to be processed and the data to be processed corresponding to the object to be processed into the language model, and obtain at least one promotion description information corresponding to the object to be processed, the data to be processed Including at least one of the following: text information, picture information, video information; Wherein, the language model includes a regression estimation sub-algorithm and a rule inference sub-algorithm, and the regression estimation sub-algorithm is used to analyze the data elements included in the data to be processed and the relationship between the various data elements to obtain the element relationship Result: The rule inference sub-algorithm is used to establish a language rule system, and the rule system is used to analyze and process the element relationship result to determine the output data of the language model. 13. An electronic device, comprising: a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein when the one or more computer instructions are executed by the processor The data processing method as claimed in claim 11 is realized.

Images