A personalized social copywriting system and method driven by AI

Abstract

The present invention discloses a personalized social copywriting system and method based on AI drive, which belongs to the field of information processing technology, wherein the system includes: a copywriting data acquisition module; a copywriting clustering module, which is used to extract a copywriting style feature vector from a copywriting sample, and cluster according to the copywriting style feature vector; a copywriting type module; a copywriting template module, which is used to establish a corresponding relationship between the matched copywriting parameters and the positions of the copywriting keywords in the copywriting sample to obtain a copywriting template; a copywriting parameter acquisition module; a copywriting style selection module; a copywriting generation module, which is used to randomly select a copywriting template according to the popularity according to the social platform, copywriting type and copywriting style, and generate a copywriting based on the copywriting template and copywriting parameters; and a copywriting quality evaluation module. The present invention can quickly generate social copywriting according to user needs.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to an AI-driven personalized social copywriting system and method. Background Art

[0002] Merchants need to provide corresponding copywriting when promoting and obtaining product experience information from customers. The effect of the copywriting affects the publicity and promotion effect of the product or activity. However, if merchants want to ensure the quality of the copywriting, they need professional writers to write copywriting templates, which places high demands on writers. In the related art, the corresponding copywriting is generated by applying a preset copywriting template. This method solves the problem of automatic copywriting generation, but it is impossible to determine the publicity effect of the copywriting and update the copywriting template in time according to the publicity effect. In addition, the preset copywriting templates often carry the writer's own style, and the diversity of the copywriting style cannot be guaranteed, so as to use the copywriting style that is of interest to different groups of people in a targeted manner. Therefore, there is an urgent need for a method that can generate corresponding high-quality copywriting based on the characteristics of different groups of people and different platform styles to ensure the effectiveness of publicity and promotion. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an AI-driven personalized social copywriting system and method, which can quickly generate social copywriting according to user needs.

[0004] On the one hand, an embodiment of the present invention provides an AI-driven personalized social copywriting system, including: a copywriting data collection module, which is used to retrieve and obtain copywriting from different social platforms according to product keywords, and filter copywriting samples according to popularity and store them in a copywriting database; a copywriting clustering module, which is used to extract copywriting style feature vectors from copywriting samples, and cluster them according to the copywriting style feature vectors to obtain a first clustering result; assign copywriting style labels to copywriting samples according to the clustering results; and extract product description language feature vectors from copywriting samples, and cluster them according to the copywriting product description language feature vectors to obtain a second clustering result; and calculate the copywriting style feature vector according to the first clustering result and the second clustering result. The influence coefficient on the popularity of the copy sample; the copy type module is used to classify the copy purpose, obtain the copy type, and set the corresponding copy parameters for each copy type; the copy types include: favorable copy, event invitation, customer service, return visit to customers, event recommendation and product recommendation; the copy parameters include: greeting, product name, product description, event name, event time and store address; the copy template module is used to assign a number of copy samples to each copy type, extract copy keywords from the copy samples, match the copy parameters according to the copy keywords, establish a corresponding relationship between the matched copy parameters and the position of the copy keywords in the copy samples, obtain the copy template and store it to the copy database, and associate the copy template with the popularity, copy style, copy type and social platform corresponding to the copy sample; determine whether to add or delete the copy template according to the copy parameters required by the copy type; a copy parameter acquisition module is used to automatically obtain copy parameter information according to the copy parameter type, and match the corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; the copy parameter keywords corresponding to the copy parameter information include: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age; a copy style selection module, Used to determine the copy style according to the user's copy style selection operation in the interactive interface, or match the corresponding copy style according to customer characteristic parameters; the customer characteristic parameters include customer gender and age; a copy generation module, used to determine the social platform, copy type, copy style and copy parameters in response to the user's operation in the interactive interface, randomly select a copy template according to the influence coefficient of the copy style according to the social platform, the copy type and the copy style, and generate the copy based on the copy template and the copy parameters; a copy quality evaluation module, used to track the feedback results of the copy within a set time after the copy is published to the social platform or sent to the customer's social account, and determine the copy quality according to the feedback results.

[0005] An AI-driven personalized social copywriting system according to an embodiment of the present invention includes at least the following beneficial effects: an embodiment of the present invention generates corresponding copywriting templates by acquiring copywriting samples with high popularity on different platforms to enrich copywriting diversity. And assign several copywriting templates to different copywriting types, and assign copywriting parameters to different copywriting types to suit different application scenarios. Copywriting templates of different styles are obtained by copywriting clustering, so as to provide different styles of copywriting for different types of people in a targeted manner to improve the copywriting publicity effect. And cluster the copywriting samples according to the product description, obtain the influence of various product descriptions and various copywriting styles on the popularity of the copywriting samples, and select the corresponding copywriting style according to the size of the influence coefficient; and save the user's parameter setting operation steps by automatically obtaining copywriting parameter information. The quality of the copywriting is ensured by tracking the feedback results of the copywriting and conducting a copywriting quality assessment.

[0006] On the other hand, an embodiment of the present invention provides a personalized social copywriting method driven by AI, comprising: S100, retrieving and acquiring copywriting from different social platforms according to product keywords, and filtering copywriting samples according to popularity and storing them in a copywriting database; S200, extracting copywriting style feature vectors from copywriting samples, and clustering them according to the copywriting style feature vectors to obtain a first clustering result; assigning copywriting style labels to copywriting samples according to the clustering result; and extracting product description feature vectors from copywriting samples, and clustering them according to the copywriting product description feature vectors to obtain a second clustering result; and calculating the copywriting style feature vector according to the first clustering result and the second clustering result. The influence coefficient of the quantity on the popularity of the copy sample; S300, classify the purpose of the copy to obtain the copy type, and set the corresponding copy parameters for each copy type; the copy types include: favorable copy, event invitation, customer service, return visit to customers, event recommendation and product recommendation; the copy parameters include: greeting, product name, product description, event name, event time and store address; S400, assign a number of copy samples to each copy type, extract copy keywords from the copy samples, match the copy parameters according to the copy keywords, establish a corresponding relationship between the matched copy parameters and the position of the copy keywords in the copy samples, and obtain the copy template And store it in the copy database, and associate the copy template with the popularity, copy style, copy type and social platform corresponding to the copy sample; determine whether to add or delete the copy template according to the copy parameters required by the copy type; S500, automatically obtain copy parameter information according to the copy parameter type, and match the corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; the copy parameter keywords corresponding to the copy parameter information include: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age; S60 0. Determine the copy style according to the copy style selection operation of the user in the interactive interface, or match the corresponding copy style according to the customer characteristic parameters; the customer characteristic parameters include the customer's gender and age; S700. In response to the user's operation in the interactive interface, determine the social platform, copy type, copy style and copy parameters, randomly select a copy template according to the influence coefficient of the copy style according to the social platform, the copy type and the copy style, and generate the copy based on the copy template and the copy parameters; S800. Track the feedback results of the copy within a set time after the copy is published to the social platform or sent to the customer's social account, and determine the copy quality according to the feedback results.

[0007] An AI-driven personalized social copywriting method according to an embodiment of the present invention includes at least the following beneficial effects: an embodiment of the present invention generates corresponding copywriting templates by acquiring copywriting samples with high popularity on different platforms to enrich copywriting diversity. And assign several copywriting templates to different copywriting types, and assign copywriting parameters to different copywriting types to suit different application scenarios. Copywriting templates of different styles are obtained by copywriting clustering, so as to provide different styles of copywriting for different types of people in a targeted manner to improve the copywriting publicity effect. And the copywriting samples are clustered according to the product descriptions to obtain the influence of various product descriptions and various copywriting styles on the popularity of the copywriting samples, and the corresponding copywriting style is selected according to the size of the influence coefficient; and the user's parameter setting operation steps are saved by automatically obtaining copywriting parameter information. The quality of the copywriting is ensured by tracking the feedback results of the copywriting and conducting a copywriting quality assessment.

[0008] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0010] Figure 1 A schematic block diagram of modules of a system according to an embodiment of the present invention;

[0011] Figure 2 A schematic diagram of generating a favorable comment copy according to an embodiment of the present invention;

[0012] Figure 3 This is a schematic diagram of parameter settings for creating text for a circle of friends according to an embodiment of the present invention;

[0013] Figure 4 Based on Figure 3 A schematic diagram of a text generated by the text parameters set in the embodiment;

[0014] Figure 5 A schematic diagram of setting parameters for an activity invitation text according to an embodiment of the present invention;

[0015] Figure 6 Based on Figure 5 A schematic diagram of a text generated by the text parameters set in the embodiment;

[0016] Figure 7 A schematic diagram of customer service copy parameter settings according to an embodiment of the present invention;

[0017] Figure 8 A schematic diagram of setting parameters for a customer return visit copywriting according to an embodiment of the present invention;

[0018] Fig. 9 A schematic diagram of copywriting parameter settings recommended for a free creation activity according to an embodiment of the present invention;

[0019] Fig.10 A schematic diagram of copywriting parameter settings recommended for a freely created product according to an embodiment of the present invention;

[0020] Fig.11 The figure is a flowchart of a method according to an embodiment of the present invention.

[0021] Reference numerals:

[0022] The copy data collection module 100 , the copy clustering module 200 , the copy type module 300 , the copy template module 400 , the copy parameter acquisition module 500 , the copy style selection module 600 , the copy generation module 700 , and the copy quality assessment module 800 . DETAILED DESCRIPTION

[0023] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0024] In the description of the present invention, "several" means one or more, "more" means two or more, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0025] Reference Figure 1 , an embodiment of the present invention provides an AI-driven personalized social copywriting system, including:

[0026] The copywriting data collection module 100 is used to retrieve copywriting from different social platforms according to product keywords, and filter copywriting samples according to popularity and store them in the copywriting database.

[0027] The copy clustering module 200 is used to extract the copy style feature vector from the copy sample, and cluster according to the copy style feature vector to obtain a first clustering result; assign a copy style label to the copy sample according to the clustering result of the copy clustering module; and extract the product description feature vector from the copy sample, and cluster according to the copy product description feature vector to obtain a second clustering result; and calculate the influence coefficient of the copy style feature vector on the popularity of the copy sample based on the first clustering result and the second clustering result.

[0028] The copy type module 300 is used to classify the purpose of the copy, obtain the copy type, and set corresponding copy parameters for each copy type; the copy types include: positive copy, event invitation, customer service, return customer visit, event recommendation and product recommendation; the copy parameters include: greeting, product name, product description, event name, event time and store address.

[0029] The copy template module 400 is used to assign a number of copy samples to each copy type, extract copy keywords from the copy samples, match copy parameters according to the copy keywords, establish a correspondence between the matched copy parameters and the positions of the copy keywords in the copy samples, obtain the copy template and store it in the copy database, and associate the copy template with the popularity, copy style, copy type and social platform corresponding to the copy sample; determine whether to add or delete the copy template according to the copy parameters required by the copy type.

[0030] The copy parameter acquisition module 500 is used to automatically acquire copy parameter information according to the copy parameter type, and match the corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; matching the corresponding copy parameter keywords according to the copy parameter information includes: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age.

[0031] The copywriting style selection module 600 is used to determine the copywriting style according to the copywriting style selection operation of the user in the interactive interface, or to match the corresponding copywriting style according to the customer characteristic parameters; the customer characteristic parameters include the customer's gender and age.

[0032] The copywriting generation module 700 is used to determine the social platform, copywriting type, copywriting style and copywriting parameters in response to the user's operation in the interactive interface, randomly select a copywriting template according to the social platform, copywriting type and copywriting style according to the influence coefficient of the copywriting style, and generate the copy based on the copywriting template and copywriting parameters.

[0033] The copy quality assessment module 800 is used to track the feedback results of the copy within a set time after the copy is published on the social platform or sent to the customer's social account, and determine the quality of the copy based on the feedback results.

[0034] It is understandable that when the copywriting samples obtained from various platforms cannot be applied to all application scenarios corresponding to the copywriting types, the copywriting template module of the embodiment of the present invention can also be used to write copywriting templates by professional writers to supplement the copywriting templates required for scenarios that the copywriting samples cannot be applied to. In this case, the popularity, social platform and copywriting style corresponding to such copywriting templates are set by manual judgment.

[0035] In some embodiments, the copy data collection module 100 includes: a keyword screening unit, which is used to retrieve copy from different social platforms by product name to obtain a first copy sample; a copy matching unit, which is used to calculate the matching degree of the first copy sample by a number of product description words, and screen the copy with a matching degree greater than a matching degree threshold to obtain a second copy sample; a popularity screening unit, which is used to determine whether the number of second copy samples exceeds the copy quantity threshold, and if so, screen the second copy samples according to the popularity to obtain a third copy sample; wherein the popularity is calculated based on the number of likes, comments and reposts of the copy samples on the social platform.

[0036] In some embodiments, calculating the matching degree of the first copy sample through several product description words includes: obtaining product description words one by one from the set of product description words, determining whether there are words or synonyms identical to the product description words in the copy sample, and if so, adding the product description words to the first product description word set; adding the weights corresponding to all product description words in the first product description word set to obtain the matching degree of the copy sample.

[0037] In some embodiments, the copy clustering module 200 includes: a copy analysis unit, which is used to perform word segmentation and part-of-speech analysis on copy sentences, obtain address words, modal particles and punctuation marks in the copy, and determine the types of address words, the types and quantities of adjectives, the types and quantities of modal particles, and the types and quantities of sentence types in the copy; sentence types include exclamatory sentences, interrogative sentences and declarative sentences; a feature vector construction unit, which is used to construct a copy style feature vector according to the types of address words, the types and quantities of adjectives, the types and quantities of modal particles, and the types and quantities of sentence types; a clustering unit, which is used to cluster copy samples according to the copy style feature vector to obtain clustering results.

[0038] In some embodiments, the copy clustering module 200 includes: a customer feature matching unit, which is used to obtain copy samples that are of interest to several people of different age groups and different genders and extract a first copy style feature vector, and obtain the copy style center feature vectors of several cluster centers of the clustering results, calculate the distance between the first copy style feature vector and the copy style center feature vector, and associate the cluster center corresponding to the minimum distance of the calculation result with the age group and gender corresponding to the first copy style feature vector to obtain the customer features corresponding to the cluster center; the customer features include age group and gender.

[0039] In some embodiments, the document clustering module includes:

[0040] The second clustering unit is used to extract product descriptions from the copywriting samples, construct product description feature vectors according to the product descriptions, and perform clustering according to the product description feature vectors to obtain a second clustering result;

[0041] Influence coefficient unit, used to build heat model :

[0042]

[0043] Among them, a is the influence coefficient of product description, b is the influence coefficient of copywriting style, and c is the influence coefficient of the number of followers of the copywriting sample source account. is the description of the i-th product, is the jth copywriting style; and the influence coefficient is fitted according to the heat model.

[0044] In some embodiments, determining whether to perform addition and deletion operations on a copy template according to copy parameters required by a copy type includes: obtaining copy parameters corresponding to the copy type to obtain a first copy parameter set; obtaining copy parameters that establish a corresponding relationship with the copy template to obtain a second copy parameter set; comparing the first copy parameter set with the second copy parameter set; when any copy parameter in the first copy parameter set does not exist in the second copy parameter set, adding a sentence containing copy parameter-related content to the copy template, and establishing a corresponding relationship with the copy parameter at a position where the copy parameter-related content is located in the sentence; when any copy parameter in the second copy parameter set does not exist in the first copy parameter set, deleting the sentence at a position in the copy template where the copy parameter is located.

[0045] In some embodiments, the copy generation module 700 includes: a copy template acquisition unit, which is used to respond to user operations on the interactive interface and obtain the copy template corresponding to the social platform, copy type, and copy style selected by the user; a copy template selection unit, which is used to set a first random number range, assign a corresponding second random number range to each copy template according to the popularity corresponding to the acquired copy template, randomly generate a random number according to the first random number range, and determine the corresponding copy template according to the second random number range where the random number is located; a copy generation unit, which is used to generate copy according to copy parameters and copy templates.

[0046] In some embodiments, the copy quality assessment module 800 includes: a social platform copy quality assessment unit, which is used to obtain the number of likes, comments and reposts after a set time after the copy is posted on the social platform, and calculate the copy quality based on the number of likes, comments and reposts. The system of the embodiment of the present invention also includes: a copy heat adjustment module, which is used to obtain any two copy templates of the same copy type, obtain the absolute value of the difference between the heat ratio and the copy quality ratio of the two copy templates, and if the absolute value of the difference is greater than a threshold, the copy quality of the two copy templates is updated as a new heat.

[0047] In some embodiments, the copy quality assessment module 800 includes: a customer account copy quality assessment unit, which is used to obtain customer feedback information after a set time after the copy is sent to the customer's social account, and evaluate the copy quality based on the customer feedback information; the customer feedback information includes the customer's product purchase information or activity participation information.

[0048] Reference Figure 2 , Figure 3 and Fig. 9 In a specific embodiment, the present invention sets three types of copywriting, including favorable comments, inspiration and free creation. Inspiration and free creation include several copywriting subcategories. Among them, inspiration includes copywriting types for circle of friends creation, event invitation, customer service and return visit to customers. Free creation includes copywriting types for event recommendation and product recommendation on different platforms, including Xiaohongshu, Meituan / Dazhong, and circle of friends. Figure 2 In this embodiment, the system automatically generates favorable comments based on the store name, address and product features preset by the system. Figure 3 and Figure 4 In this embodiment, the creative text of the circle of friends is generated by setting the scene parameters and the text style. Figure 5 and Figure 6 In this embodiment, the activity invitation text is automatically generated by setting the parameter information of the activity invitation, including the activity name, activity time and activity content. Figure 7 and Figure 8 In this embodiment, the copywriting is generated by setting the scenario case information of customer service and return visit to customers. Fig. 9 and Fig.10 In this embodiment, by selecting the copywriting types of different platforms, selecting activity recommendation or product recommendation in the copywriting types, and setting corresponding parameters, the copywriting with different platform styles and conforming to the set parameters is generated according to the user's selection.

[0049] Corresponding to the above-mentioned embodiments, the present invention also provides a method embodiment. As for the method embodiment, since it basically corresponds to the system embodiment, the relevant parts can be referred to the partial description of the system embodiment.

[0050] Reference Fig.11 , an embodiment of the present invention provides a personalized social copywriting method based on AI driving, comprising the following steps:

[0051] S100, retrieve and obtain copywriting from different social platforms according to product keywords, and filter copywriting samples according to popularity and store them in a copywriting database.

[0052] S200, extracting a copywriting style feature vector from the copywriting sample, and clustering the copywriting style feature vector to obtain a first clustering result; assigning a copywriting style label to the copywriting sample according to the clustering result; extracting a product description feature vector from the copywriting sample, and clustering the copywriting product description feature vector to obtain a second clustering result; and calculating an influence coefficient of the copywriting style feature vector on the popularity of the copywriting sample according to the first clustering result and the second clustering result;

[0053] S300, classifying the purpose of the copywriting to obtain the copywriting type, and setting corresponding copywriting parameters for each copywriting type; the copywriting types include: favorable copywriting, event invitation, customer service, return visit to customers, event recommendation and product recommendation; the copywriting parameters include: greeting, product name, product description, event name, event time and store address;

[0054] S400, allocating a number of copywriting samples to each copywriting type, extracting copywriting keywords from the copywriting samples, matching copywriting parameters according to the copywriting keywords, establishing a corresponding relationship between the matched copywriting parameters and the positions of the copywriting keywords in the copywriting samples, obtaining copywriting templates and storing them in a copywriting database, and associating the copywriting templates with the popularity, copywriting style, copywriting type, and social platform corresponding to the copywriting samples; determining whether to add or delete the copywriting templates according to the copywriting parameters required by the copywriting type;

[0055] S500, automatically obtaining copy parameter information according to copy parameter type, and matching corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; matching corresponding copy parameter keywords according to copy parameter information includes: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age;

[0056] S600, determining a copywriting style according to a copywriting style selection operation performed by a user in an interactive interface, or matching a corresponding copywriting style according to customer characteristic parameters; the customer characteristic parameters include customer gender and age;

[0057] S700, in response to the user determining the social platform, the copy type, the copy style and the copy parameters in the interactive interface operation, randomly selecting a copy template according to the social platform, the copy type and the copy style according to the influence coefficient of the copy style, and generating the copy based on the copy template and the copy parameters;

[0058] S800. After the copy is published on a social platform or sent to a customer's social account, the copy feedback result is tracked within a set time, and the copy quality is determined based on the feedback result.

[0059] Although specific embodiments are described herein, those of ordinary skill in the art will recognize that many other modifications or alternative embodiments are also within the scope of the present disclosure. For example, any of the functions and / or processing capabilities described in conjunction with a particular device or component may be performed by any other device or component. In addition, although various exemplary implementations and architectures have been described according to embodiments of the present disclosure, those of ordinary skill in the art will recognize that many other modifications to the exemplary implementations and architectures described herein are also within the scope of the present disclosure.

[0060] It will be appreciated by those skilled in the art that all or some of the steps and systems in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0061] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present application. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present application.

Claims

1. An AI-driven personalized social copywriting system, characterized in that: include: The copywriting data collection module is used to retrieve copywriting from different social platforms based on product keywords, and filter copywriting samples based on popularity and store them in the copywriting database; A copywriting clustering module is used to extract a copywriting style feature vector from a copywriting sample, and cluster the copywriting style feature vector to obtain a first clustering result; assign a copywriting style label to the copywriting sample according to the clustering result; and extract a product description feature vector from the copywriting sample, and cluster the copywriting product description feature vector to obtain a second clustering result; and calculate the influence coefficient of the copywriting style feature vector on the popularity of the copywriting sample according to the first clustering result and the second clustering result; The copywriting type module is used to classify the copywriting purposes, obtain the copywriting types, and set the corresponding copywriting parameters for each copywriting type; The copy types include: favorable comments, event invitations, customer service, returning customers, event recommendations and product recommendations; the copy parameters include: greetings, product names, product descriptions, event names, event times and store addresses; The copywriting template module is used to assign a number of copywriting samples to each copywriting type, extract copywriting keywords from the copywriting samples, match copywriting parameters according to the copywriting keywords, establish a corresponding relationship between the matched copywriting parameters and the positions of the copywriting keywords in the copywriting samples, obtain the copywriting template and store it in the copywriting database, and associate the copywriting template with the popularity, copywriting style, copywriting type and social platform corresponding to the copywriting sample; determine whether to add or delete the copywriting template according to the copywriting parameters required by the copywriting type; A copy parameter acquisition module is used to automatically acquire copy parameter information according to the copy parameter type, and match corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; the copy parameter keywords corresponding to the copy parameter information include: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age; A copywriting style selection module, used to determine the copywriting style according to the copywriting style selection operation of the user in the interactive interface, or to match the corresponding copywriting style according to customer characteristic parameters; the customer characteristic parameters include customer gender and age; A copywriting generation module, configured to determine a social platform, a copywriting type, a copywriting style, and copywriting parameters in response to a user operation on an interactive interface, randomly select a copywriting template according to the social platform, the copywriting type, and the copywriting style according to an influence coefficient of the copywriting style, and generate a copywriting based on the copywriting template and the copywriting parameters; The copywriting quality assessment module is used to track the feedback results of the copywriting within a set time after the copywriting is published on the social platform or sent to the customer's social account, and determine the quality of the copywriting based on the feedback results.

2. The AI-driven personalized social copywriting system according to claim 1, characterized in that: The document data collection module includes: A keyword screening unit, used to retrieve copywriting from different social platforms by product name to obtain a first copywriting sample; A copy matching unit, configured to calculate a matching degree for the first copy sample using a plurality of product description words, and screen the copy with a matching degree greater than a matching degree threshold to obtain a second copy sample; a heat screening unit, configured to determine whether the number of the second copywriting samples exceeds a copywriting number threshold, and if so, to screen the second copywriting samples according to the heat to obtain a third copywriting sample; The popularity is calculated based on the number of likes, comments and reposts of the copywriting sample on the social platform.

3. The AI-driven personalized social copywriting system according to claim 2, characterized in that: The calculating the matching degree of the first copywriting sample by using a plurality of product description words comprises: Obtain product description words one by one from the set of product description words, determine whether there are words or synonyms identical to the product description words in the copy sample, and if so, add the product description words to the first set of product description words; The weights corresponding to all product description words in the first product description word set are added together to obtain the matching degree of the copywriting sample.

4. The AI-driven personalized social copywriting system according to claim 1, characterized in that: The document clustering module includes: A copy analysis unit, used to perform word segmentation and part-of-speech analysis on the copy sentences, obtain address words, modal particles and punctuation marks in the copy, determine the types of address words, the types and numbers of adjectives, the types and numbers of modal particles, and the types and numbers of sentence types in the copy; the sentence types include exclamatory sentences, interrogative sentences and declarative sentences; A feature vector construction unit, used to construct a copywriting style feature vector according to the types of address words, the types and quantities of adjectives, the types and quantities of modal particles, and the types and quantities of sentence types; A clustering unit, used to cluster the copywriting samples according to the copywriting style feature vector to obtain a clustering result; The customer feature matching unit is used to obtain copywriting samples that are of interest to a number of people of different age groups and different genders and extract a first copywriting style feature vector, and obtain the copywriting style center feature vectors of a number of cluster centers of the clustering results, calculate the distance between the first copywriting style feature vector and the copywriting style center feature vector, associate the cluster center corresponding to the minimum distance of the calculation result with the age group and gender corresponding to the first copywriting style feature vector, and obtain the customer features corresponding to the cluster center; the customer features include age group and gender.

5. The AI-driven personalized social copywriting system according to claim 4, characterized in that: The document clustering module includes: A second clustering unit is used to extract product descriptions from the copywriting samples, construct a product description feature vector according to the product descriptions, and perform clustering according to the product description feature vector to obtain a second clustering result; Influence coefficient unit, used to construct the heat model H: H=aP i +bS j +cf Among them, a is the influence coefficient of product description, b is the influence coefficient of copywriting style, f is the number of followers of the copywriting sample source account, c is the influence coefficient of the number of followers of the copywriting sample source account, P i is the description of the i-th product, S j is the jth copywriting style; an influence coefficient is fitted according to the heat model, and the influence coefficient includes: an influence coefficient of the product description, an influence coefficient of the copywriting style, and an influence coefficient of the number of followers of the copywriting sample source account.

6. The AI-driven personalized social copywriting system according to claim 1, characterized in that: Determining whether to add or delete a copy template according to the copy parameters required by the copy type includes: Obtain the text parameters corresponding to the text type to obtain the first text parameter set; Acquire the text parameters that establish a corresponding relationship with the text template to obtain a second text parameter set; Comparing the first copy parameter set with the second copy parameter set; When any one of the text parameters in the first text parameter set does not exist in the second text parameter set, a sentence containing text parameter-related content is added to the text template, and a corresponding relationship is established between the text parameter and the position of the text parameter-related content in the sentence; When any one of the text parameters in the second text parameter set does not exist in the first text parameter set, the sentence at the position corresponding to the text parameter in the text template is deleted.

7. The AI-driven personalized social copywriting system according to claim 1, characterized in that: The copywriting generation module includes: A copy template acquisition unit, used to obtain a copy template corresponding to the social platform, copy type, and copy style selected by the user in response to the user's operation on the interactive interface; a text template selection unit, configured to set a first random number range, assign a corresponding second random number range to each text template according to an influence coefficient of a text style corresponding to the acquired text template, randomly generate a random number according to the first random number range, and determine a corresponding text template according to the second random number range in which the random number is located; The copy generation unit is used to generate copy according to copy parameters and copy templates.

8. The AI-driven personalized social copywriting system according to claim 1, characterized in that: The copywriting quality assessment module includes: A social platform copywriting quality evaluation unit, used to obtain the number of likes, comments and reposts of the copywriting after a set time after it is posted on the social platform, and calculate the copywriting quality based on the number of likes, comments and reposts; The system further comprises: The copy heat adjustment module is used to obtain any two copy templates of the same copy type, obtain the absolute value of the difference between the heat ratio and the copy quality ratio of the two copy templates, and if the absolute value of the difference is greater than a threshold, update the copy quality of the two copy templates as the new heat.

9. The AI-driven personalized social copywriting system according to claim 1, characterized in that: The copywriting quality assessment module includes: A customer account copy quality assessment unit is used to obtain customer feedback information after a set time after the copy is sent to the customer's social account, and assess the copy quality based on the customer feedback information; the customer feedback information includes the customer's product purchase information or activity participation information.

10. A personalized social copywriting method based on AI, characterized in that: The following steps are involved: S100, retrieve and obtain copywriting from different social platforms according to product keywords, and filter copywriting samples according to popularity and store them in a copywriting database; S200, extracting a copywriting style feature vector from the copywriting sample, and clustering the copywriting style feature vector to obtain a first clustering result; assigning a copywriting style label to the copywriting sample according to the clustering result; extracting a product description feature vector from the copywriting sample, and clustering the copywriting product description feature vector to obtain a second clustering result; and calculating an influence coefficient of the copywriting style feature vector on the popularity of the copywriting sample according to the first clustering result and the second clustering result; S300, classifying the purpose of the copy to obtain the copy type, and setting corresponding copy parameters for each copy type; The copy types include: favorable comments, event invitations, customer service, returning customers, event recommendations and product recommendations; the copy parameters include: greetings, product names, product descriptions, event names, event times and store addresses; S400, allocating a number of copywriting samples to each copywriting type, extracting copywriting keywords from the copywriting samples, matching copywriting parameters according to the copywriting keywords, establishing a corresponding relationship between the matched copywriting parameters and the positions of the copywriting keywords in the copywriting samples, obtaining copywriting templates and storing them in a copywriting database, and associating the copywriting templates with the popularity, copywriting style, copywriting type, and social platform corresponding to the copywriting samples; determining whether to add or delete the copywriting templates according to the copywriting parameters required by the copywriting type; S500, automatically acquiring copy parameter information according to copy parameter type, and matching corresponding copy parameter keywords according to the copy parameter information; the copy parameter type includes time, date, customer gender and customer age; matching corresponding copy parameter keywords according to the copy parameter information includes: matching morning, noon and evening greetings according to time, matching holiday greetings according to date, and matching customer address words according to customer gender and customer age; S600, determining a copywriting style according to a copywriting style selection operation performed by a user in an interactive interface, or matching a corresponding copywriting style according to customer characteristic parameters; the customer characteristic parameters include customer gender and age; S700, in response to a user operating in an interactive interface to determine a social platform, a copy type, a copy style, and copy parameters, randomly selecting a copy template according to the social platform, the copy type, and the copy style according to an influence coefficient of the copy style, and generating a copy based on the copy template and the copy parameters; S800. After the copy is published on a social platform or sent to a customer's social account, the copy feedback result is tracked within a set time, and the copy quality is determined based on the feedback result.

Images