Text generation method, electronic device, and computer-readable storage medium

Abstract

This application discloses a text generation method, electronic device, and computer-readable storage medium, relating to the fields of artificial intelligence and text generation. The method comprises: obtaining a structured description text of a target object and an example text corresponding to the structured description text, wherein the example text is used to describe an object of the same type as the target object; inputting the structured description text and the example text into a first text generation model, and using the first text generation model to generate multiple initial texts corresponding to the target object, wherein the multiple initial texts have the same attributes as the example text; placing the multiple initial texts into a preset scenario, and collecting feedback information about the multiple initial texts; and generating a target text for the target object using a second text generation model based on the feedback information and the multiple initial texts. This application solves the technical problem of poor text generation performance in related technologies.

Description

Technical Field

[0001] The present application relates to the fields of artificial intelligence and text generation, and more specifically, to a text generation method, an electronic device, and a computer-readable storage medium. Background Art

[0002] In recent years, large-scale language models have developed rapidly, and their application in intelligent text generation has significantly improved the efficiency and quality of text creation. However, the disconnect between text generation and actual delivery has become a major challenge. In traditional processes, text generation, testing, selection, and final delivery take a long time, and lack an effective feedback mechanism to guide model improvement. As a result, the generated text may not accurately match user preferences, resulting in poor text generation results.

[0003] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0004] The embodiments of the present application provide a text generation method, an electronic device, and a computer-readable storage medium to at least solve the technical problem of poor text generation effect in related technologies.

[0005] According to one aspect of an embodiment of the present application, a text generation method is provided, including: obtaining a structured description text of a target object and an example text corresponding to the structured description text, wherein the example text is used to describe an object of the same type as the target object; inputting the structured description text and the example text into a first text generation model, and using the first text generation model to generate multiple initial texts corresponding to the target object, wherein the multiple initial texts have the same attributes as the example text; placing the multiple initial texts into a preset scene, and collecting feedback information of the multiple initial texts, wherein the feedback information is used to indicate the degree of user preference for the multiple initial texts in the preset scene; and generating a target text of the target object using a second text generation model based on the feedback information and the multiple initial texts, wherein the target text conforms to the feedback information.

[0006] According to one aspect of an embodiment of the present application, a text generation method is also provided, including: obtaining a structured description text of a target product and an example title text corresponding to the structured description text, wherein the example title text is used to describe a product of the same type as the target product; inputting the structured description text and the example title text into a first text generation model, and using the first text generation model to generate multiple initial title texts corresponding to the target product, wherein the multiple initial title texts have the same attributes as the example title text; placing the multiple initial title texts into a preset scene, and collecting feedback information of the multiple initial title texts, wherein the feedback information is used to indicate the degree of user preference for the multiple initial title texts in the preset scene; and using a second text generation model to generate a target title text for the target product based on the feedback information and the multiple initial title texts, wherein the target title text conforms to the feedback information.

[0007] According to one aspect of an embodiment of the present application, a text generation method is also provided, including: responding to an input instruction on an operation interface, displaying a structured description text of a target object and an example text corresponding to the structured description text on the operation interface, wherein the example text is used to describe an object of the same type as the target object; responding to a processing instruction on the operation interface, displaying a target text of the target object on the operation interface, wherein the target text is obtained by using a second text generation model based on feedback information and multiple initial texts, the feedback information is feedback information of multiple initial texts collected by putting multiple initial texts into a preset scene, the multiple initial texts are obtained by inputting the structured description text and the example text into a first text generation model, the multiple initial texts have the same attributes as the example text, and the target text conforms to the feedback information.

[0008] According to one aspect of an embodiment of the present application, a text generation method is also provided, including: obtaining a structured description text of a target object and an example text corresponding to the structured description text by calling a first interface, wherein the first interface includes a first parameter, and the parameter value of the first parameter includes the structured description text and the example text; obtaining the structured description text of the target object and the example text corresponding to the structured description text, wherein the example text is used to describe an object of the same type as the target object; inputting the structured description text and the example text into a first text generation model, and using the first text generation model to generate multiple initial texts corresponding to the target object, wherein the multiple initial texts have the same attributes as the example text; placing the multiple initial texts into a preset scene, and collecting feedback information of the multiple initial texts, wherein the feedback information is used to indicate the user's preference for the multiple initial texts in the preset scene; based on the feedback information and the multiple initial texts, using a second text generation model to generate a target text of the target object, wherein the target text conforms to the feedback information; outputting the target text by calling a second interface, wherein the second interface includes a second parameter, and the parameter value of the second parameter includes the target text.

[0009] According to another aspect of the embodiments of the present application, a computing device is further provided, including: a memory storing an executable program; and a processor for running the program, wherein the method of each embodiment of the present application is executed when the program is running.

[0010] According to another aspect of the embodiments of the present application, an electronic device is also provided, including: a memory storing an executable program; a processor connected to the memory via a bus, and configured to run the program, wherein the method of each embodiment of the present application is executed when the program is running.

[0011] According to another aspect of an embodiment of the present application, a computer-readable storage medium is also provided, which includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the methods in various embodiments of the present application.

[0012] According to another aspect of the embodiments of the present application, a computer program product is further provided, including a computer program, which implements the methods in various embodiments of the present application when executed by a processor.

[0013] According to another aspect of an embodiment of the present application, a computer program product is also provided, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method in each embodiment of the present application is implemented.

[0014] According to another aspect of the embodiments of the present application, a computer program is further provided, which implements the methods in various embodiments of the present application when executed by a processor.

[0015] In an embodiment of the present application, a structured description text of a target object and an example text corresponding to the structured description text are obtained, wherein the example text is used to describe an object of the same type as the target object; the structured description text and the example text are input into a first text generation model, and a plurality of initial texts corresponding to the target object are generated using the first text generation model, wherein the plurality of initial texts have the same attributes as the example text; the plurality of initial texts are placed in a preset scene, and feedback information of the plurality of initial texts is collected, wherein the feedback information is used to indicate the degree of user preference for the plurality of initial texts in the preset scene; based on the feedback information and the plurality of initial texts, a target text of the target object is generated using a second text generation model, wherein the target text conforms to the feedback information, thereby achieving the purpose of improving the text generation effect. It is easy to notice that a plurality of initial texts with the same attributes as the sample text can be generated by using the first text generation model through the sample text that describes the same type of object as the target object. The plurality of initial texts can be put into the preset scene, and the initial text features with higher user preference in the preset scene can be identified by analyzing the feedback information, and the second text generation model can be used to adjust its generation strategy so that the generated target text is more likely to have these features, while reducing or avoiding the initial text features with lower user preference in the target text, thereby improving the pertinence and user acceptance of text generation, and thus solving the technical problem of poor text generation effect in related technologies.

[0016] It is easy to notice that the above general description and the following detailed description are merely for the purpose of exemplifying and explaining the present application, and do not constitute a limitation to the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a schematic diagram of an application scenario according to an embodiment of the present application;

[0019] Figure 2 is a flowchart of a text generation method according to an embodiment of the present application;

[0020] Figure 3 This is a flowchart of product information structure in an intelligent copywriting LLM according to an embodiment of the present application;

[0021] Figure 4is a flow chart of structured information output according to an embodiment of the present application;

[0022] Figure 5 This is a flow chart of a large language model generated by a copywriting according to an embodiment of the present application;

[0023] Figure 6 is a schematic diagram of a process for constructing preference training data according to an embodiment of the present application;

[0024] Figure 7 is a schematic diagram of a model training process according to an embodiment of the present application;

[0025] Figure 8 is a flowchart of a text generation method according to an embodiment of the present application;

[0026] Figure 9 is a flowchart of a text generation method according to an embodiment of the present application;

[0027] Figure 10 is a flowchart of a text generation method according to an embodiment of the present application;

[0028] Figure 11 is a schematic diagram of a text generation device according to an embodiment of the present application;

[0029] Figure 12 is a schematic diagram of a text generation device according to an embodiment of the present application;

[0030] Figure 13 is a schematic diagram of a text generation device according to an embodiment of the present application;

[0031] Figure 14 is a schematic diagram of a text generation device according to an embodiment of the present application;

[0032] Figure 15 is a structural block diagram of a computing device according to an embodiment of the present application;

[0033] Figure 16 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] The technical solution provided in this application is primarily implemented using a content generation model, which refers to a deep learning model with large-scale model parameters. This model, also known as a foundation model, is pre-trained on large-scale, unlabeled corpora to produce pre-trained models with over 100 million parameters. This model is adaptable to a wide range of downstream tasks and exhibits good generalization capabilities, such as large language models (LLMs) and multi-modal pre-training models.

[0037] It should be noted that when the content generation model is actually applied, the pre-trained model can be fine-tuned through a small number of samples, so that the content generation model can be applied to different tasks. For example, the content generation model can be widely used in natural language processing (NLP), computer vision, speech processing and other fields. Specifically, it can be applied to computer vision tasks such as visual question answering (VQA), image description (IC), image generation, etc. It can also be widely used in natural language processing tasks such as text-based sentiment classification, text summary generation, and machine translation. Therefore, the main application scenarios of the content generation model include but are not limited to digital assistants, intelligent robots, search, online education, office software, e-commerce, intelligent design, etc. In the embodiments of the present application, data processing through the content generation model in the text generation scenario is used as an example for explanation.

[0038] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following interpretations:

[0039] Generative optimization: In the advertising creative delivery process, a variety of candidate advertising creative materials are first generated, and then the materials with good delivery effects are selected through a carousel method, and the poor materials are eliminated.

[0040] Optimal Generation: By accumulating data on random intelligent creative placement over a long period of time using small batches of products, combined with multi-dimensional information such as placement scenario summaries, merchant information, and store crowds, preference alignment algorithms such as Direct Preference Optimization (DPO) and Proximal Policy Optimization (PPO) are used to achieve optimal generation of intelligent creative materials (LLM). This breaks away from the "cast a wide net" generation and replacement model and significantly improves the return on investment (ROI) of production capacity.

[0041] Retrieval-Augmented Generation (RAG): refers to adjusting the output of large language models so that they can reference knowledge bases other than the training data source before generating responses.

[0042] Preference alignment: Preference alignment technology aims to allow machines to continuously improve their behavior by receiving human feedback, so as to better meet human expectations.

[0043] Direct Preference Optimization (DPO): A method for improving language models based on human preferences. Unlike reinforcement learning from human feedback (RLHF), DPO does not rely on explicit reward modeling or reinforcement learning. It targets the same goal as RLHF but provides a simpler and more direct training method.

[0044] Currently, the integrated intelligent copy generation and delivery project has achieved steady growth in the revenue per thousand impressions (RPM) of display ads through improved headlines. Traditional intelligent copy (and other creative assets) iterations rely on delivery performance, but the process from production, publishing, cold-start carousel, online selection, and final delivery takes at least a week. This, coupled with the "cast a wide net" approach to generative optimization, leads to slow iteration and low productivity. LLM technology is rapidly developing, and content generation models can easily generate some high-quality copy. However, these copywritings do not always achieve positive results in actual delivery. This is because copywriting generation and delivery are decoupled, lacking a feedback mechanism from delivery to the generation model, resulting in uncertain positive results.

[0045] This application abandons the traditional generative optimization and proposes an improved generation paradigm. The innovations include: 1. Proposing a structured LLM for product information to achieve controllable generation of selling points, such as the number of words; 2. Proposing a RAG-based text generation LLM to improve the diversity of intelligent texts; 3. Proposing an improved generation LLM, based on actual delivery effects, using a preference alignment algorithm to generate intelligent texts with relatively good delivery effects.

[0046] According to an embodiment of the present application, a text generation method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0047] Considering the huge amount of model parameters of the content generation model and the limited computing resources of the mobile terminal, the above method provided in the embodiment of the present application can be applied to Figure 1 The application scenarios shown are not limited to these. Figure 1 is a schematic diagram of an application scenario according to an embodiment of the present application. Figure 1 In the illustrated application scenario, the content generation model is deployed on a server 10. The server 10 can be connected to one or more client devices 20 via a local area network (LAN), a wide area network (WAN), the Internet, or other types of data networks. The client devices 20 herein may include, but are not limited to, smartphones, tablet computers, laptops, PDAs, personal computers, smart home devices, and in-vehicle devices. The client devices 20 can interact with users via a graphical user interface (GUI) to invoke the content generation model and thereby implement the methods provided in the embodiments of the present application.

[0048] In an embodiment of the present application, a system consisting of a client device and a server may perform the following steps: the client device generates a structured description text and a sample text corresponding to the structured description text. The server inputs the structured description text and the sample text into a first text generation model, and generates multiple initial texts corresponding to a target object using the first text generation model; places the multiple initial texts into a preset scenario, and collects feedback information about the multiple initial texts; and based on the feedback information and the multiple initial texts, generates a target text for the target object using a second text generation model.

[0049] It should be noted that, with the rapid development of high-performance computing units, in other application scenarios, the above method provided in the embodiment of the present application can also be applied to the model all-in-one machine. In an optional embodiment, a variety of models are built into the model all-in-one machine, and the user can choose to adjust with a model as needed to obtain the user's own model, so that the high-performance computing unit built into the model all-in-one machine can directly call the adjusted model to execute the above method provided in the embodiment of the present application. In another optional embodiment, a trained model is built into the content generation model all-in-one machine, so that the high-performance computing unit built into the model all-in-one machine can directly call the model to execute the above method provided in the embodiment of the present application.

[0050] Furthermore, when users need to train their own models, they can upload their own datasets through the client. This dataset is then sent to the server, which then adjusts the pre-trained model using the dataset to create the user's own model, which can then be deployed in production. To facilitate user model adjustment needs, the server provides a complete set of adjustment tools, development frameworks, and processes, supporting a variety of adjustment strategies, making the adjusted model more adaptable to different application fields and highly customized.

[0051] Under the above operating environment, this application provides Figure 2 The text generation method shown. Figure 2 Flowchart of the text generation method according to the embodiment of the present application. Figure 2 As shown, the method may include the following steps:

[0052] Step S202: obtaining a structured description text of the target object and an example text corresponding to the structured description text;

[0053] The sample text is used to describe an object of the same type as the target object.

[0054] The target objects mentioned above refer to the products or services for which advertising copy needs to be created, and are not limited here.

[0055] The above-mentioned structured description text can be a kind of refined and formatted information, which presents the key attributes of the product (such as brand, style, and function) in a standard form, making it easier for the algorithm to understand and process it, and reducing errors and illusions in the generation process.

[0056] The above-mentioned sample texts refer to advertising copy samples that have been effective in the past. These copy samples demonstrate the effectiveness of attributes and expressions for similar products, providing references and applications for the generation of new copy.

[0057] In an optional embodiment, the original information of the target product is first collected, including the product title, pictures, videos, attributes, etc., and then a specific algorithm is used to process this information, extract key points, and form a structured description text. In addition, the system can call a sample text library consisting of high-quality copywriting that matches the target product type. This step can ensure the accuracy and consistency of the input information, providing high-quality raw materials for subsequent intelligent copywriting generation. At the same time, by introducing sample text, the system can learn effective copywriting attributes and expressions, thereby improving the starting level of copywriting generation.

[0058] Through structured information processing and guidance from sample texts, the diversity and quality of copywriting can be significantly improved, reducing the problem of hallucinations in the model generation process, ensuring that the generated copywriting more accurately reflects the characteristics of the product and is more attractive to target users in terms of attributes.

[0059] Step S204: inputting the structured description text and the sample text into a first text generation model, and using the first text generation model to generate a plurality of initial texts corresponding to the target object;

[0060] Among them, the properties of the multiple initial texts are the same as those of the sample texts.

[0061] The first text generation model mentioned above can be an intelligent copywriting LLM, which is a trained large-scale language model that can understand the input product structured description text and example text and generate diverse but consistent copywriting.

[0062] The above-mentioned structured description text may refer to converting the original information of the product (such as title, picture, video, attributes, etc.) into a descriptive text in a standard format through multimodal information processing technology, which is convenient for model understanding and generation.

[0063] The above-mentioned sample text can be a sample of copywriting with excellent results in the past, which is used to generate copywriting with similar attributes using the model.

[0064] The above-mentioned multiple initial texts can be a series of preliminary copywriting generated by the intelligent copywriting creation LLM based on the structured description text and sample text. While the attributes of these copywritings are consistent with the sample texts, the content is adjusted for the target product.

[0065] In an optional embodiment, the system inputs the structured description text of the target object and the sample text that matches its type into the intelligent copywriting LLM. By understanding these input texts, the model uses its internal generation mechanism and retrieval-augmented generation technology (RAG) to generate multiple initial texts with the same attributes as the sample text. Specifically, RAG technology allows the model to retrieve and refer to high-quality text instances related to the target object when generating copy, ensuring that the output copy not only covers the core concerns of the object, but also remains consistent with the sample text in terms of expression and attributes, thereby improving the diversity and quality of the text while reducing the hallucination problems that may be encountered during the generation process. Through the generation capabilities of the intelligent copywriting LLM and the retrieval function of RAG technology, the diversity and quality of the text are significantly improved, while maintaining consistency with the attributes of the sample text.

[0066] Step S206: placing the multiple initial texts into a preset scene and collecting feedback information of the multiple initial texts;

[0067] The feedback information is used to indicate the user's preference for multiple initial texts in a preset scenario.

[0068] The above-mentioned preset scenario can be a specific display environment of the target object. For example, it can refer to a specific display scenario of an advertisement, but it is not limited to this. The display scenario can refer to when a user visits an e-commerce platform for the first time or searches for a product, and the system guesses and displays product advertisements that may attract the user's attention based on their past behavior, preferences, and search keywords. The display scenario can also refer to the system displaying advertisements for related or complementary products to the purchased product after the user completes the purchase to promote secondary purchases or improve user satisfaction. The display scenario can also be an advertising display scenario for non-core traffic or non-peak hours, such as inconspicuous locations in recommendation streams, advertising displays for low-activity user groups, etc.

[0069] The aforementioned feedback information refers to user responses to the delivered content, including but not limited to metrics such as click-through rate, conversion rate, and user dwell time. This data can reflect how the content performs in the intended scenario and user preferences for the content. By analyzing this feedback information, we can determine which content is most effective in attracting users, thereby improving advertising effectiveness.

[0070] The aforementioned user preference refers to the user's preference for the initial text, which can often be indirectly measured through metrics such as click-through rate (CTR) in feedback. A high CTR often indicates user interest in the text content, and therefore, indicates a high level of user preference for the text in the intended scenario. Other metrics in feedback, such as conversion rate, can also further reveal whether the text effectively guides users to complete purchases, thereby comprehensively understanding user preferences.

[0071] In an optional embodiment, multiple initial texts (ad titles or descriptions) can be delivered to different pre-set scenarios. These scenarios can include a variety of advertising scenarios, each with different user traffic and behavior patterns. The system records the specific performance of each text in different scenarios, such as clicks, impressions, and conversions. Based on this data, the click-through rate and conversion rate of each text can be calculated.

[0072] By delivering multiple initial texts in pre-set scenarios and collecting feedback, the system can identify the text types and expressions that users prefer. This not only increases the appeal of ads but also improves overall advertising effectiveness, including higher user engagement, click-through rates, and conversion rates. Furthermore, through continuous delivery and feedback loops, the system can continuously refine the text generation model, making the generated text more in line with user preferences, thereby achieving continuous improvement in advertising effectiveness.

[0073] Step S208 : generating a target text of the target object using a second text generation model based on the feedback information and the multiple initial texts.

[0074] Among them, the target text conforms to the feedback information.

[0075] In this solution, the second text generation model mentioned above specifically refers to the intelligent copy optimization model (LLM). It is a trained model that generates optimal copy based on historical campaign performance and user preferences. The model's role at this stage is to integrate feedback and adjust its generation strategy to ensure that the output copy is more accurately aligned with user preferences, thereby improving market effectiveness.

[0076] The target text described above is the text used for delivery. It is the result of improvements based on user feedback based on the initial text. This text not only contains key product information, but also better aligns with the target audience's preferences in terms of presentation and attributes, potentially increasing click-through and conversion rates.

[0077] In an optional embodiment, the system first analyzes feedback collected from pre-set scenarios to identify text features that perform well, including attributes, information structure, and expression methods. These features are then used as utilization signals and input into a second text generation model. Based on these feedback signals, the model adjusts its generation strategy to produce better target text. Optionally, the model may be trained using a Descriptive Proof-of-Purpose (DPO) algorithm to improve its strategy and ensure that the output text is more likely to be clicked and converted by users.

[0078] By improving the secondary text generation model based on user feedback, we can significantly improve the delivery performance of generated text. The target text's attributes and expression are more aligned with user preferences, increasing click-through and conversion rates and attracting users. This closed-loop improvement mechanism closely links text generation with delivery performance, enabling continuous iteration and improvement of text generation, ensuring that delivered text is a highly refined, screened, and improved version.

[0079] The above steps describe the core process of the integrated intelligent copy generation and delivery solution, from product information processing to improving delivery text.

[0080] In an optional embodiment, the system extracts structured description text of the target object (such as a product), which is a systematic summary of the product's features, attributes, and selling points for model understanding and generation. At the same time, the system also collects sample texts, which represent descriptions of objects of the same type as the target object. Usually, the sample text has a positive effect on delivery and can guide the generation of text with consistent attributes. After obtaining the structured description text and sample text, the system inputs this information into the first text generation model. Based on the attributes of the sample text, the model generates multiple initial texts related to the target object. These texts are consistent with the sample text in attributes, and are combined with specific information of the target object to form a diverse text sample.

[0081] Next, the system delivers these initial texts to pre-set scenarios to test user preference. Pre-set scenarios are selected based on traffic and target user group analysis, with the goal of gathering feedback on their effectiveness. Through delivery, the system collects feedback on multiple initial texts. This information reflects user responses to the texts in pre-set scenarios, such as click-through rates and conversion rates, providing direct data on user preference.

[0082] Finally, based on the collected feedback and the initial text, the system uses a secondary text generation model to generate target text. This target text is an improved description that better aligns with user preferences and is expected to achieve better market results in the intended scenario. Through this closed-loop process, the system can continuously iterate and improve its text generation strategy, ensuring that each target text delivered is market-proven, engaging, and highly effective.

[0083] The above steps complete a closed loop from intelligent product information processing to improved text delivery, enhancing the quality and market responsiveness of generated text. The introduction of structured descriptive text reduces the problem of illusory text generation and improves information accuracy. The use of sample text guides the consistency of text attributes and increases its appeal. The collection and analysis of feedback provides a data foundation for subsequent text improvements, ensuring that the generated target text is more aligned with user preferences, thereby improving click-through and conversion rates.

[0084] Through the above steps, a structured description text of the target object and an example text corresponding to the structured description text are obtained, wherein the example text is used to describe an object of the same type as the target object; the structured description text and the example text are input into a first text generation model, and a plurality of initial texts corresponding to the target object are generated by using the first text generation model, wherein the attributes of the plurality of initial texts are the same as those of the example text; the plurality of initial texts are put into a preset scene, and feedback information of the plurality of initial texts is collected, wherein the feedback information is used to indicate the degree of user preference for the plurality of initial texts in the preset scene; based on the feedback information and the plurality of initial texts, a target text of the target object is generated by using a second text generation model, wherein the target text meets the feedback information. information, thereby achieving the purpose of improving the text generation effect; it is easy to notice that, through the sample text that describes the same type of object as the target object, multiple initial texts with the same attributes as the sample text can be generated by the first text generation model, and the multiple initial texts can be put into the preset scene. By analyzing the feedback information, the initial text features with a higher degree of user preference in the preset scene are identified, and the second text generation model is used to adjust its generation strategy, so that the generated target text is more likely to have these features, while reducing or avoiding the initial text features with a lower degree of user preference from appearing in the target text, thereby improving the pertinence of text generation and user acceptance, and thus solving the technical problem of poor text generation effect in related technologies.

[0085] In the above embodiment of the present application, based on the feedback information and multiple initial texts, a target text of a target object is generated using a second text generation model, including: based on the feedback information, selecting a first text and a second text from the multiple initial texts, wherein the user preference degree of the first text is greater than the user preference degree of the initial texts other than the first text, and the user preference degree of the second text is less than the user preference degree of the initial texts other than the second text; based on the first text and the second text, generating the target text using the second text generation model.

[0086] The above-mentioned feedback information refers to the user's response feedback to multiple initial texts in specific preset scenarios, which is used to quantify the user's preference for the text, such as user click-through rate, dwell time, conversion rate and other data indicators.

[0087] The first and second texts mentioned above are representative samples selected from multiple initial texts, where the user preference for the first text is relatively high, while the user preference for the second text is relatively low. They are used to guide the training of subsequent models to generate target texts that are more in line with user preferences.

[0088] The above-mentioned second text generation model can generate LLM for intelligent copywriting optimization. A trained model can generate better target text based on existing feedback information and text samples to improve user participation and response effects in preset scenarios.

[0089] The target text mentioned above refers to the descriptive text generated by the second text generation model (Intelligent Copywriting Optimization Generation LLM), which is guided and improved by preference information. It is used to more efficiently display it to users in preset scenarios, in order to obtain higher user preference and market response.

[0090] In an optional embodiment, the system generates a series of diverse initial texts using the intelligent copywriting model (LLM) based on structured descriptions and sample text. These texts are then randomly distributed in pre-set scenarios to collect user feedback on the different texts, including click-through rates, conversion rates, and other information to quantify user preferences. After collecting this feedback, the system selects a first text (i.e., a text with a relatively high user preference) and a second text (a text with a relatively low user preference) from the multiple initial texts. This selection process essentially involves the system ranking the initial texts based on feedback data and using these differences in preferences to guide subsequent model training.

[0091] Based on the preference differences between the first and second texts, the system uses the second text generation model to learn and adjust to generate a target text that is more aligned with the user's preferences. This guidance process can be achieved using a preference alignment algorithm (such as DPO). By comparing and analyzing user preferences for different texts, the model can learn how to generate text that is more appealing to the target user.

[0092] For example, let's assume there's a newly released pair of children's sneakers. The system first generates multiple initial text descriptions based on the product description and sample text, including Initial Text 1, Initial Text 2, and Initial Text 3. These texts are then placed in different display ad scenarios, and the system collects user feedback on each description. Suppose Initial Text 2 achieves a relatively high click-through rate and conversion rate, while Initial Text 3 has a relatively low click-through rate and conversion rate. In this case, Initial Text 2 becomes the first text, and Initial Text 3 becomes the second text.

[0093] Based on the preference differences between the first and second texts, the system uses the second text generation model to learn how to adjust the generation strategy to produce a target text that is more appealing to users. For example, the system might use the model to pay more attention to descriptions of brand identity, seasonality, comfort, and sporty attributes because these elements are emphasized in the first text and are verified by higher user preference.

[0094] The above steps, through a closed-loop intelligent generation mechanism, can improve the market responsiveness of text. In practical applications, this process can accelerate the iteration speed of intelligently generated text, reduce resource waste during the generation process, and ensure that the generated text is more accurately aligned with user preferences, thereby improving delivery effectiveness.

[0095] In the above embodiment of the present application, based on the first text and the second text, the target text is generated using the second text generation model, including: obtaining the scene description text of the preset scene and the associated description text of the target object; inputting the scene description text, the associated description text, the first text and the second text into the second text generation model, and generating the target text using the second text generation model.

[0096] The above scenario description text is the specific environment or conditions when the text is delivered. The user traffic characteristics and behavior patterns in different scenarios are different.

[0097] The above-mentioned associated description text can be additional descriptive information related to the target object. For example, the associated description text can be a product information description, such as the detailed attributes, selling points and price of the product, etc. This is not limited to this. The associated description text can also be other types of descriptive text associated with the target object.

[0098] In an optional embodiment, the system must first determine the preset scene in which the target object is located, and collect scene descriptions and associated description texts related to the scene and the product. The scene description provides the context for the placement, while the associated description text covers the characteristics of the merchant and consumer, as well as information about the product itself. After collecting the necessary information, the system provides the scene description text, associated description text, as well as the first text with a higher user preference and the second text with a lower user preference as input to the second text generation model. This model analyzes these inputs, especially under the guidance of the scene description and the first text, to learn how to generate target text that is more attractive to the target user group, while avoiding the poorly performing features in the second text.

[0099] Furthermore, based on this input, the second text generation model generates target text that better aligns with the context of the pre-set scenario and the information in the target object's associated description. Furthermore, thanks to model learning and adjustments, the expression and information presentation are more tailored to user needs, adapting to user preferences. This process completes a closed-loop improvement process, from product information to market verification, and then to model training and text generation.

[0100] By combining scene descriptions and related description texts, as well as user feedback, the second text generation model is used to generate target text that is more in line with market and user preferences, thereby improving the attractiveness and conversion efficiency of the copy.

[0101] In the above embodiment of the present application, a first text and a second text are selected from multiple initial texts based on feedback information, including: in response to the feedback information being the amount of user interaction in a preset scenario, sorting the multiple initial texts based on the amount of user interaction to obtain a sorting result; matching the multiple initial texts based on the sorting result to obtain multiple text pairs, wherein the degree of text difference in different text pairs is different; screening the multiple text pairs based on the degree of text difference to obtain a target text pair, wherein the degree of text difference of the target text pair is greater than the degree of text difference of other text pairs in the multiple text pairs except the target text pair, and the target text pair includes the first text and the second text.

[0102] The above ranking results are the relative rankings of the initial texts in terms of user preference obtained by analyzing and ranking the user interaction amounts.

[0103] The above text pairs are two texts selected as a group from the sorted initial texts for subsequent text difference and preference comparison analysis.

[0104] The above-mentioned text difference degree refers to the difference in expression, information content or attributes between the first text and the second text in the text pair, and is used to evaluate the diversity of the text generated by the model.

[0105] The target text pair is a pair of texts selected from multiple text pairs, which has relatively large text differences and includes texts with relatively high and relatively low user preferences, and is used for training and adjusting the second text generation model.

[0106] In an optional embodiment, the system sorts multiple initial texts based on the amount of user interaction in the feedback information (such as click-through rate, conversion rate, etc.) to obtain a sorting result, which reflects the user's preference for different texts. The goal of sorting is to identify texts that users prefer and dislike. Next, the system selects paired texts (i.e., text pairs) from the sorted multiple initial texts for comparison, and the degree of difference between these text pairs varies. The purpose of constructing text pairs is to subsequently analyze the impact of text differences on user preferences. On the basis of establishing multiple text pairs, the system needs to screen out target text pairs by analyzing the degree of text difference of each text pair. The target text pair here refers to a pair with a relatively large degree of text difference among all text pairs. The target text pair contains a first text with a relatively high degree of user preference and a second text with a relatively low degree of preference, which is used for subsequent training and adjustment of the guidance model.

[0107] Determining the target text pair described above helps the second text generation model learn and adjust. Based on the information from the first text (higher preference) and the second text (lower preference) in the target text pair, the model will generate a target text that better matches the user's preferences. For example, if text A emphasizes brand and sport characteristics, while text B focuses on price and promotional information, the model may learn that in advertising scenarios, users prefer descriptions that include brand and sport characteristics, and thus place greater emphasis on these elements when generating the target text.

[0108] The above steps significantly improve the quality and diversity of text generation through a closed-loop feedback mechanism and differential analysis. Specifically, by sorting the initial text based on the amount of user interaction in the feedback information and screening target text pairs with relatively large text differences, the system can more accurately identify and understand user preferences, guiding the intelligent copywriting optimization method (LLM) to generate more attractive descriptive text. This mechanism not only accelerates the text generation and optimization process and reduces unnecessary resource consumption, but also ensures that the generated text is more accurately aligned with user preferences, improving key indicators such as click-through rate and conversion rate.

[0109] In the above embodiment of the present application, the method also includes: obtaining multiple sample texts of the sample object; placing the multiple sample texts into multiple sample scenarios, and collecting sample feedback information of the multiple sample texts, wherein the sample feedback information is used to represent the user preference degree for the multiple sample texts in the multiple sample scenarios; selecting a first sample text and a second sample text from the multiple sample texts based on the sample feedback information, wherein the user preference degree of the first sample text is greater than the user preference degree of the sample texts other than the first sample text, and the user preference degree of the second sample text is less than the user preference degree of the sample texts other than the second sample text; generating a target sample text using an initial text generation model based on the first sample text and the second sample text; adjusting the model parameters of the initial text generation model based on the first sample text, the second sample text and the target sample text to obtain a second text generation model.

[0110] The sample object mentioned above may refer to a product or service instance for which a description text needs to be generated.

[0111] The aforementioned multiple sample texts may be a series of text sets for describing sample objects, generated by the initial text generation model based on the structured information of the product or service.

[0112] The above-mentioned sample scenarios refer to preset environments or conditions for delivering and testing sample texts.

[0113] The sample feedback information mentioned above is collected from users by placing sample texts in sample scenarios. It is used to measure user preference for different texts, including key metrics such as click-through rate and conversion rate. The first and second sample texts are selected from multiple sample texts, representing sample texts with high and low user preference, respectively, and are used to guide and adjust model parameters.

[0114] The target sample text is generated by the initial text generation model and adjusted based on the preference information of the first sample text and the second sample text.

[0115] The above-mentioned model parameter adjustment is to modify the parameters of the initial text generation model based on the comparative analysis of the first sample text, the second sample text and the target sample text, so as to generate text that better meets the user preferences and scenario requirements.

[0116] In an optional embodiment, the system generates multiple sample texts based on the structured information of the sample objects using an initial text generation model. These texts describe different aspects of the sample objects. The sample texts are then placed in multiple sample scenarios to collect user feedback in different scenarios and assess user preference for the sample texts.

[0117] The system analyzes collected sample feedback (such as click-through rate and conversion rate) and ranks sample texts based on user preference, selecting a first sample text (highly preferred by users) and a second sample text (lower preferred by users). This selection process essentially involves a quantitative analysis of user feedback, as well as the ranking and classification of text effectiveness. Based on the preference information for the first and second sample texts, the system utilizes an initial text generation model to generate target sample text. This process involves adjusting the model generation strategy to produce text that is more appealing to users and better suited to scenario requirements.

[0118] Finally, based on a comparative analysis of the first sample text, the second sample text, and the target sample text, the system adjusts the parameters of the initial text generation model. This adjustment is typically performed using a preference alignment algorithm to enable the model to generate text that better meets user preferences and scenario requirements, thereby obtaining the second text generation model.

[0119] The above steps continuously iterate and improve the text generation model through feedback from actual delivery results to generate higher-quality descriptive text. Specifically, by testing and analyzing user feedback information from multiple sample texts in multiple sample scenarios, the system can identify which description methods are more popular with users and which perform poorly. Based on these analysis results, the system uses the model to generate target sample text that is closer to user preferences. By adjusting the model parameters, it continuously improves the model's generation strategy to ensure that the generated text is more accurately aligned with user preferences and improve key indicators such as click-through rate and conversion rate in the scenario.

[0120] In the above embodiment of the present application, the model parameters of the initial text generation model are adjusted based on the first sample text, the second sample text and the target sample text to obtain the second text generation model, including: constructing a first loss function based on the first sample text and the target sample text, wherein the first loss function is used to increase the similarity between the target sample text and the first sample text; constructing a second loss function based on the second sample text and the target sample text, wherein the second loss function is used to reduce the similarity between the target sample text and the second sample text; adjusting the model parameters based on the first loss function and the second loss function to obtain the second text generation model.

[0121] The first sample text mentioned above refers to a text sample that is identified as having a relatively high user preference through actual delivery and user feedback information.

[0122] The target sample text is generated by the initial text generation model and adjusted based on the preference information of the first sample text and the second sample text, aiming to be a text that is closer to user preferences and scenario requirements.

[0123] The second sample text mentioned above is opposite to the first sample text. The second sample text is a text sample with a relatively low user preference, and is used for comparison and adjustment of model parameters.

[0124] The first loss function is used to increase the similarity between the target sample text and the first sample text (text with high user preference) to take advantage of the model's tendency to generate text that is more popular with users.

[0125] The second loss function is used to reduce the similarity between the target sample text and the second sample text (text with low user preference) so as to use the model to avoid generating text that the user does not like.

[0126] In an optional embodiment, the system constructs a first loss function based on the first sample text and the target sample text, aiming to increase the similarity between the two. Simultaneously, a second loss function is constructed based on the second sample text and the target sample text, aiming to reduce the similarity between the target sample text and the second sample text (text with low user preference). By constructing these two types of loss functions, the system can guide the adjustment of model parameters from both positive and negative perspectives.

[0127] After constructing the loss function, the system adjusts the model parameters of the initial text generation model based on the guidance of the first and second loss functions. This adjustment process typically involves training and iterating the machine learning algorithm. By reducing the similarity with low-preference text and increasing the similarity with high-preference text, the model learns how to adjust its generation strategy to output text that better suits user preferences.

[0128] By continuously adjusting the parameters of the initial text generation model, we eventually obtain a second text generation model trained with a preference alignment algorithm. This model can more accurately generate text that is highly preferred by users when describing products, reducing the cost of generating and eliminating inefficient text and improving conversion rates.

[0129] Through a closed-loop feedback mechanism and guided by a loss function, the above steps improve the quality and market responsiveness of intelligently generated copy, reduce the cost of generating and eliminating inefficient copy, and increase conversion rates. Specifically, by increasing the similarity with the first sample text (high user preference text) and decreasing the similarity with the second sample text (low user preference text), the system can leverage the initial text generation model to generate more accurate and engaging copy. This not only reduces the resource waste of generating large amounts of inefficient copy, but also improves click-through rates, conversion rates, and user engagement in the scenario, thereby achieving greater advertising effectiveness and market responsiveness.

[0130] In the above embodiment of the present application, based on the first sample text and the second sample text, the target sample text is generated using the initial text generation model, including: obtaining sample scene description texts of multiple sample scenes and sample association description texts of sample objects; inputting the sample scene description texts, the sample association description texts, the first sample text and the second sample text into the initial text generation model, and generating the target sample text using the initial text generation model.

[0131] The above-mentioned sample scenario description text provides the environment or condition information of the sample preset scenario, such as user traffic, purchase intention, etc., so as to use the model to generate copy that is more in line with the characteristics of the scenario.

[0132] The above-mentioned sample-related description text is information text closely related to the sample object. For example, the sample-related description text can include detailed attributes, selling points, brand information, price, etc., as well as information such as merchant reputation, consumer purchasing power and preferences, which is used to enhance the personalization and accuracy of the model-generated copy.

[0133] In an optional embodiment, the system first collects sample scene description texts of multiple sample scenes and sample associated description texts of sample objects. This information covers the scene characteristics and product characteristics, as well as the background information of merchants and consumers, and constitutes a comprehensive information basis for guiding the model to generate copy. The above-collected sample scene description texts, sample associated description texts, first sample texts (with a relatively high degree of user preference) and second sample texts (with a relatively low degree of user preference) are input into the initial text generation model. The system uses this information to generate new target sample texts using the model. The input of multi-dimensional information can ensure that the copy generated by the model is more personalized and fits the scene requirements. At the same time, by comparing the first sample text and the second sample text, the model can learn the expression method and information presentation that are more popular with users.

[0134] By combining multi-dimensional information input, the initial text generation model is guided to generate more accurate and effective copy. Specifically, the system collects and analyzes user feedback on multiple sample texts in different sample scenarios, and determines the first sample text with a relatively high user preference and the second sample text with a relatively low user preference. Subsequently, based on the scene description, product information, merchant and consumer background information, the system uses the model to generate target sample text that is closer to user preferences and scene requirements. This process is actually a closed-loop improvement based on user feedback information, which can make the generated copy more accurately attract the target user group, improve click-through rate, conversion rate and user engagement, and thus achieve higher advertising effectiveness and market responsiveness.

[0135] By combining multi-dimensional information input, the personalization and scenario adaptability of the copy are significantly enhanced, and the user's willingness to click and purchase is increased. Specifically, by inputting information such as sample scenario description text and sample associated description text into the initial text generation model, the system can use the model to generate copy that is more in line with user needs and scenario characteristics. This not only increases the attractiveness of the copy, but also ensures that the copy can more accurately convey the characteristics of the product and stimulate users' purchasing interest. This mechanism not only reduces the waste of resources in the generative optimization process through closed-loop feedback and the integration of multi-dimensional information, but also continuously iterates and improves the text generation strategy according to the dynamic changes in user preferences and scenario needs, ensuring the market effect and personalization level of the copy.

[0136] In the above embodiment of the present application, obtaining a structured description text of a target object and an example text corresponding to the structured description text includes: obtaining multimodal data of the target object, wherein the multimodal data is used to describe the target object through different modalities; converting the multimodal data into an initial description text of the target object; performing structured processing on the initial description text to obtain a structured description text; and retrieving an example text from a database based on the type of the target object, wherein the database is used to store example texts of objects of different types.

[0137] Multimodal data refers to data about a target object collected through various information modes (e.g., text, images, video, audio, etc.). Structured descriptive text: This textual information, converted and processed from multimodal data, contains clear, organized key information about the target object, such as product attributes, brand, and features.

[0138] The above-mentioned initial description text is text information extracted directly from multimodal data without structured processing. It is usually messy or lengthy and requires further processing before it can be used as model input.

[0139] The above database stores a collection of example texts of different types of objects. These example texts can be verified and effective copywriting or descriptions, which are used to guide and inspire the model to generate higher quality text.

[0140] The above example texts are representative text samples of specific types of objects, which are used to enhance the model generation capabilities. By learning the expression and information structure of these example texts, the model can generate copy that is more in line with scenario requirements and user preferences.

[0141] In an optional embodiment, the system first obtains multimodal data of the target object, which may include multiple information sources such as product titles, attribute descriptions, pictures, videos, etc., forming a comprehensive description of the target object. Subsequently, the system converts the collected multimodal data into initial description text. This process may involve technologies such as optical character recognition, image annotation, and video annotation to ensure the integrity and accuracy of the text information. The obtained initial description text is processed through key information summary and structured LLM model to extract and organize key information of the product, such as brand, appearance, ingredients, efficacy and function, etc., to obtain structured description text. This step converts lengthy and messy text information into a clear, organized structured form to facilitate model understanding and generate more accurate copy. Finally, based on the type of target object, the system retrieves relevant sample text from the database. These sample texts provide a reference to help the model learn and imitate effective copy attributes and information presentation methods to generate higher quality and more user-attractive text.

[0142] The above steps significantly enhance the model's input information and generation capabilities through the fusion processing and structured transformation of multimodal data, reduce the LLM illusion problem, and improve the accuracy and quality of copy generation. Specifically, by collecting multimodal data on products and converting and structuring it into clear, organized, structured description text, the system can provide the model with more accurate and comprehensive product information, thereby generating copy that is more in line with product characteristics and user preferences. In addition, by retrieving example text of specific types of objects in the database, the model can learn a wider and more diverse range of copy attributes, improving the diversity and attractiveness of the generated copy, and demonstrating the innovation and practicality of the technical solution in the field of e-commerce advertising copy generation.

[0143] Through structured information input and an efficient retrieval mechanism, the LLM model significantly reduces the problem of hallucinations when generating copy, improving the accuracy and quality of the copy. Specifically, structured processing ensures that the model receives clear and organized product information, thus avoiding the occurrence of statements that do not match the actual product during the generation process, and improving the accuracy and credibility of the copy. At the same time, the mechanism for retrieving example text provides the model with a rich learning resource, helping it generate more attractive and diverse copy, further increasing user click-through and purchase intentions, and demonstrating the practical value and market potential of this technical solution in the field of e-commerce advertising copy generation.

[0144] This application includes three core algorithm modules: a large language model for intelligent copywriting creation, a large language model for intelligent copywriting optimization generation, and project efficiency improvement and customer expression. For the large language model for intelligent copywriting creation, the open source LLM can be adjusted through structured input of product information, imitation of intelligent copywriting attributes based on RAG, and building a data flywheel by collecting high-quality copywriting from users / merchants, thereby improving the generation quality and diversity of the intelligent copywriting LLM and reducing the illusion problem. For the large language model for intelligent copywriting optimization generation, the data of random intelligent copywriting placement can be accumulated over a long period of time through small batches of products. Combined with multi-dimensional information such as placement scenario summary, merchant information, and store crowds, preference alignment algorithms such as DPO / PPO can be used to realize intelligent copywriting optimization generation LLM, get rid of the "wide-net" generation and replacement model, and significantly improve production capacity ROI. For project efficiency improvement and customer expression, the main title of the advertiser is used as the entry point to achieve project efficiency improvement in different scenarios, and complete the tool for copywriting optimization generation capabilities, thereby improving the customer usage and adoption rate of copywriting generation and recommendation tools.

[0145] Figure 3 This is a flowchart for structuring product information within the intelligent copywriting LLM, according to an embodiment of this application. This flowchart aims to improve and control copywriting generation, ensuring that it accurately reflects product characteristics while meeting the specific requirements of advertising. The diagram includes several key structures and information flows, illustrating the conversion process from raw product information to structured, controllable, and intelligent copywriting.

[0146] Figure 3 It includes three core algorithm modules: a large language model for intelligent copywriting creation, a large language model for intelligent copywriting optimization generation, and project efficiency improvement and customer expression. In the large language model for intelligent copywriting creation, a library of excellent title examples can be obtained through manual design, manual mining, and data flywheel methods. This library of excellent title examples and product structured information can be input into the large language model for intelligent copywriting creation to obtain intelligent creative titles. Various creative titles, such as merchant product titles (for example, main titles, organic traffic titles, etc.), can be published as intelligent materials. In the large language model for intelligent copywriting optimization generation, intelligent materials are input into a random carousel experiment bucket, and the scene statistical effects are analyzed to obtain click-through rate sorting (CTR) results. Information such as products, merchants, users, and delivery scenarios, as well as click-through rate sorting results, can be input into the large language model for intelligent copywriting optimization generation. The large language model for intelligent copywriting optimization generation is then used to process and obtain high-quality copy for specific product delivery scenarios. In project efficiency improvement and customer communication, we can obtain mainstream large-scale delivery and recommended imitation examples of new language attributes based on the high-quality copy of the product in the specific delivery scenario.

[0147] The data flywheel described above refers to a continuous cycle of data collection, analysis, and reapplication, designed to continuously improve the performance of a model or algorithm. The randomized rotating experiment bucket described above is a common A / B testing strategy in online advertising or recommendation systems, used to evaluate the effectiveness of different creatives or recommended content. A small portion of traffic or impressions can be randomly allocated to a set of experimental smart copywriting to observe and measure their performance in real-world environments.

[0148] For the aforementioned intelligent copywriting large language model, when the input consists only of the main product title and lacks complete product information, there are problems such as insufficient diversity in the generated results and serious hallucination problems. When the input also includes the main product title and text information extracted from visual materials, the product information is lengthy and messy, without standard format classification, resulting in problems such as lengthy input information, overly complex model tasks, inaccurate acquisition of core product information, and serious hallucination problems. This application can summarize and generalize various messy and lengthy original product information based on a multimodal model, and output clean and semantically structured information.

[0149] Figure 4 is a flow chart of structured information output according to an embodiment of the present application, such as Figure 4 As shown, it includes input information, an algorithm module, and structured output of key information. The input information can be divided into four categories: the first category is text such as titles and attributes; the second category is detail page images, where optical character recognition (OCR) can be used to convert text in images into editable and searchable text; the third category is product images, where captions can be used to automatically generate text describing the image content; and the fourth category is product videos, where video captions can be used to automatically generate text describing the video content. The algorithm module is implemented through key information summarization and a structured large language model. The structured output of key information can be a structured output of key product information.

[0150] Without fine-tuning the large language model, the problem of hallucination when the model generates titles can be effectively alleviated by referencing the structured product information. For example, the original input information causes the model to generate an incorrect brand name. After introducing structured information, the brand name and other information can be correctly extracted.

[0151] The "hallucination problem" in headline generation using large language models (LLMs) typically refers to the possibility that the model generates text inconsistent with the input information, such as incorrect brand names, non-existent product features, or irrelevant information about the product. This problem can be caused by factors such as bias in the model's training data, inaccurate understanding of the model's context, and randomness in the generation process.

[0152] Structuring product information can effectively mitigate hallucinations and accurately extract brand names. It clarifies the input format and provides a clear and consistent input format, enabling the model to more accurately locate and understand product attributes, including brand names, when processing data. Structured data is typically organized by categories, such as "product name," "brand," and "efficacy." This helps the model learn the associations and distinctions between different attributes. It also reduces noise. Raw input information may contain a large amount of irrelevant or disorganized data, which can interfere with the model's generation process. Structuring filters out irrelevant information, retaining only product features that are critical for title generation, such as brand, appearance, and material. This helps the model focus on important information and reduces generation errors. It also enhances signal strength. When the model receives a clearer and more targeted signal, it can more accurately capture specific product attributes, such as brand. Brand names are often explicitly labeled in structured data, providing the model with clear instructions to prioritize this information when generating titles, thereby avoiding generating incorrect brand names.

[0153] Structured information allows for the construction of higher-quality training datasets. These datasets better reflect the true attributes and characteristics of products, enabling the model to better understand and retain brand information during the learning process, thereby more accurately invoking and utilizing this information when generating titles. Structured information helps the model better understand the product context, namely its position in the market, target consumer group, and product characteristics. This understanding helps the model generate titles that closely align with the product's attributes, including the correct brand name.

[0154] The introduction of structured information not only reduces model processing complexity but also improves the accuracy and relevance of model-generated content, effectively mitigating the problem of hallucinations and increasing the accuracy of brand names. In specific applications, structured information clearly defines the various dimensions of a product, enabling the model to directly reference this information when generating copy. This avoids the fuzzy extraction of key information like brand names from unstructured text, significantly reducing the likelihood of generation errors.

[0155] Based on the search-enhanced copywriting generation large language model, efficient generation of various styles of title copywriting is the key to optimal generation. The method of generating specific styles of title copywriting through prompt words is relatively inefficient when adding new copywriting types. When adding new title types, the prompt words need to be redesigned and the model trained. This application can abstractly upgrade the prompt word (prompt) control generation to a general imitation model. The model can be trained only once. Subsequently, it only needs to accumulate different title copywriting examples. By retrieving example copywriting of similar products through RAG, new titles of corresponding attribute types can be produced without retraining the model, which greatly improves efficiency.

[0156] Figure 5 This is a flow chart of a large language model generated by a copywriting according to an embodiment of the present application. Figure 5 As shown, product structured information is fed into the intelligent copywriting large language model as input, providing a detailed description and key features of the product. Preferred example copy is also fed into the model as input, providing copy examples of various attributes and types. The model then uses a search-enhanced method to retrieve example copy similar to the current product in order to mimic its attributes and expressions. The intelligent copywriting large language model generates new headline copy with consistent attributes by learning from product structured information and mimicking examples from the preferred example copy library. The generated headline copy is then used in actual advertising, and its appeal and effectiveness are evaluated through real market feedback, resulting in the preferred example copy.

[0157] The aforementioned structured product information represents a structured representation of detailed product attributes, including key information such as brand, appearance, material, and function. This information is organized in a unified format, making it clear and easy to process. The intelligent copywriting large language model is trained to understand and imitate the writing attributes and structure of input examples to generate new copywriting content. Using a retrieval-enhanced approach, information from external knowledge bases can be retrieved and referenced during the generation process, improving the accuracy and diversity of the generated copy.

[0158] The aforementioned preferred sample copy library is a collection of high-quality copy, including examples of headline copy of various attributes and types. These examples can be successful advertising copy, high-quality customized headlines, or excellent copy from other sources, used to leverage and inspire the creation of new copy. Implementation involves applying the generated headlines to actual advertising campaigns to test their effectiveness and improve results.

[0159] Figure 5This paper outlines the entire process from product information to generating and improving headline copy. Its core lies in linking structured product information with a library of preferred example copywriting using RAG technology, enabling the intelligent copywriting large language model to generate more accurate copywriting that meets specific attributes. Specifically, the model first receives structured product information, including detailed product attributes. It then uses RAG technology to search the library of preferred example copywriting for copywriting examples that match the current product attributes. The model then mimics the attributes and structure of these examples to generate new headline copywriting that not only incorporates key product information but also possesses similar engaging writing attributes as the examples. The generated headline copywriting is then used in actual advertising campaigns, where its effectiveness is verified through market feedback. If it performs well, it may be added to the library of preferred example copywriting, providing more high-quality reference styles for future copywriting generation. This entire process forms a cycle that continuously accumulates and improves the model's generation capabilities, enabling diverse and efficient headline generation without the need for repeated model training or adjustments.

[0160] After a small-scale carousel experiment and accumulating effect data, the intelligent copywriting large language model aligns the effects of the generated model based on CTR ranking.

[0161] The training data involved in this application can be composed of three parts: instructions, preferred samples (Chosen), and rejected samples (Rejected). Instruction data contains specific guidance information, usually including detailed information about the product (such as product title, structured information, merchant information, etc.) and a description of the specific delivery scenario. In the Chinese context, it can be understood as "guidance instructions," providing the model with contextual information on how to generate copy. Chosen refers to copy samples that have been screened from historical delivery and have relatively good results. These copies usually have high CTR (click-through rate) or conversion rate, and are therefore selected as positive learning materials for the model, helping the model understand which copies are more likely to attract the attention of the target audience in specific scenarios. Rejected data, in contrast to Chosen samples, contains copy samples that have poor delivery performance under the same product and scenario conditions. Although these samples may seem negative, they are equally important because they can help the model learn to avoid generating similar inefficient copy.

[0162] The construction of preference training data aims to use three types of information: instruction, choice, and rejection. This allows the intelligent copy optimization model (LLM) to learn how to generate more attractive and effective copy for different products and delivery scenarios. This training method, by comparing positive and negative examples, can improve the model's copy generation quality, ensuring that it not only accurately reflects product information but also takes into account the preferences of the target audience, thereby achieving better advertising results in real-world applications.

[0163] Figure 6 is a schematic diagram of a process for constructing preference training data according to an embodiment of the present application, such as Figure 6 As shown, the construction of preference training data is a closed-loop, dynamically evolving process. Through comparative learning, the model aims to enable the intelligent copywriting optimization-based large language model to generate higher-quality copywriting that better meets the needs of the delivery scenario. The model first receives descriptions containing product information and delivery scenarios, such as the delivery scenario, merchant level descriptions, store demographics (descriptions of consumer purchasing power), and product structured information. Training is performed using a machine learning training method (Direct Preference Optimization, or DPO). High-confidence sample pairs are filtered from small-scale delivery data based on various criteria (page views, click-through rate, and conversion rate). The model then analyzes the selected and eliminated samples. Through comparative learning, it identifies which copywriting is most popular and which is least effective. This process enables the model to "learn" the characteristics that distinguish good from bad copywriting, such as language attributes, information presentation, and emotional tone. Based on the results of comparative learning, the model refines its generation strategy to improve the click-through rate and conversion rate of the output copy. When generating new copy, the model considers the product and scenario information provided in the instructions, while also drawing on the strengths of the preferred samples and avoiding the flaws of the eliminated samples, generating copy that better aligns with the target audience's preferences. The generated copy is then deployed, and its effectiveness is evaluated through market feedback. This performance data is used to update the sample library of preferred and eliminated samples, forming a data flywheel and continuously improving the model generation strategy.

[0164] Through this process, the intelligent copywriting optimization-based content generation model can continuously iterate, improving its copywriting generation capabilities for specific products and scenarios. This enables the commercialization of the copywriting optimization-based generation paradigm, significantly improving the effectiveness of copywriting delivery, and reducing the inefficiency of generative optimization. This training method enhances the model's "preference alignment" capabilities by comparing positive and negative examples, allowing it to more accurately reflect the needs and preferences of the target audience when generating copy.

[0165] For the training of large language models for intelligent copywriting optimization, we can first obtain a fusion model that can generate diverse headlines. This fusion model is used as a reference model. At the same time, the strategy model is initialized using the parameters of the fusion model. The direct preference loss (DPO Loss) is used to improve the output preference of the strategy model, making the probability of outputting positive samples higher.

[0166] Figure 7 This is a schematic diagram of a model training process according to an embodiment of the present application. Figure 7 As shown in the figure, the fusion model integrates various product information and scenario requirements to generate diverse headlines and text. The fusion model's ability to generate diverse content is crucial for subsequent preference alignment. Eliminated and selected samples can be fed into the trained fusion model (reference model) and policy model. The trained fusion model serves as a baseline for improving the policy model and comparing its output. The policy model initializes its parameters from the fusion model (reference model) and is improved using a direct preference loss to learn to generate headlines that are more preferred by users.

[0167] Through the above process, we can effectively refine the preferences of model-generated copy, making it more closely aligned with users' actual preferences, thereby achieving better results in actual advertising. By directly incorporating user feedback into the model training process, this approach avoids the long-term iteration cycles and inefficient "cast a wide net" generation methods of traditional generative optimization, significantly improving the quality of copy generation and the efficiency of advertising delivery.

[0168] Compared with the original product titles, the titles generated by the DPO preference model can put some important information of the product in the front, such as core selling points, product endorsements, and brands, and the expression is more attractive to users.

[0169] Through the construction of a data flywheel, the introduction of product structured information to reduce the illusion of a large language model, the intelligent copywriting large language model with controllable generation of multiple dimensions such as the number of selling points, and the optimization-based generation of a large language model for delivery effects, the integration of intelligent copywriting and delivery is achieved. This application proposes a structured large language model for product information to achieve controllable generation of selling points, such as the number of words; proposes a copywriting LLM based on RAG to improve the diversity of intelligent copywriting; and proposes a optimization-based generation LLM, which uses a preference alignment algorithm based on actual delivery effects to generate intelligent copywriting with relatively good delivery effects.

[0170] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0171] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

[0172] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus the necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present application.

[0173] Figure 8 is a flowchart of a text generation method according to an embodiment of the present application, such as Figure 8 As shown, the method may include the following steps:

[0174] Step S802: Obtain a structured description text of the target product and an example title text corresponding to the structured description text;

[0175] The example title text is used to describe products of the same type as the target product.

[0176] The system collects and processes detailed information about the target product, converting it into structured description text. Simultaneously, the system retrieves example title texts that match the target product type from a historical database. These example titles represent known copywriting attributes and content structures that attract user attention.

[0177] Step S804: input the structured description text and the sample title text into a first text generation model, and use the first text generation model to generate a plurality of initial title texts corresponding to the target product;

[0178] Among them, the properties of the multiple initial title texts are the same as those of the sample title texts.

[0179] The system inputs the structured description text and sample title text into a pre-trained first text generation model, and uses the model to generate a series of initial title texts with similar attributes to the sample title text. These texts contain key information about the target product and try to imitate the attractiveness and expression of the sample title.

[0180] Step S806: placing the multiple initial title texts into a preset scene and collecting feedback information of the multiple initial title texts;

[0181] The feedback information is used to indicate the user's preference for multiple initial title texts in a preset scenario.

[0182] The initial title text generated above will be placed in the preset scenario, and the system will collect user feedback information on these texts, including click-through rate, dwell time, etc., to quantify the user's preference for different attributes and content.

[0183] Step S808: Based on the feedback information and the multiple initial title texts, a target title text of the target product is generated using a second text generation model.

[0184] Among them, the target title text conforms to the feedback information.

[0185] Based on the feedback from the initial title text, the system uses a secondary text generation model to generate the final target title text. This refinement process ensures that the target title text is more aligned with user preferences, effectively attracting clicks and attention in pre-defined scenarios, and improving the text's market responsiveness.

[0186] By providing accurate product information through structured descriptive text, using example title text to guide text generation attributes, and using user feedback to evaluate and improve copywriting, the resulting title text can more accurately align with user preferences and scenario requirements, improving conversion rates and user engagement. Through this process, the system is able to generate higher-quality product title copy that is more in line with market preferences. This not only increases the attractiveness of the copy but also reduces the cost of generating and selecting inefficient copy, demonstrating the innovativeness and commercial value of the technical solution in the field of e-commerce copywriting. Through a closed-loop improvement mechanism and the integration of multi-dimensional information, the system significantly improves the quality and efficiency of intelligently generated text, providing merchants with more efficient and accurate text generation strategies.

[0187] Through the above steps, a structured description text of the target product and an example title text corresponding to the structured description text are obtained, wherein the example title text is used to describe a product of the same type as the target product; the structured description text and the example title text are input into a first text generation model, and a plurality of initial title texts corresponding to the target product are generated by the first text generation model, wherein the attributes of the plurality of initial title texts are the same as those of the example title text; the plurality of initial title texts are placed in a preset scenario, and feedback information of the plurality of initial title texts is collected, wherein the feedback information is used to indicate the degree of user preference for the plurality of initial title texts in the preset scenario; based on the feedback information and the plurality of initial title texts, a target title text of the target product is generated by a second text generation model, wherein , the target title text conforms to the feedback information, and the purpose of improving the text generation effect is achieved; it is easy to notice that, through the sample text that describes the same type of object as the target object, the first text generation model can be used to generate multiple initial texts with the same attributes as the sample text, and the multiple initial texts can be put into the preset scene. By analyzing the feedback information, the initial text features with a higher degree of user preference in the preset scene are identified, and the second text generation model is used to adjust its generation strategy, so that the generated target text is more likely to have these features, and at the same time, the initial text features with a lower degree of user preference are reduced or avoided from appearing in the target text, thereby improving the pertinence and user acceptance of text generation, and thus solving the technical problem of poor text generation effect in related technologies.

[0188] Figure 9 is a flowchart of a text generation method according to an embodiment of the present application, such as Figure 9 As shown, the method may include the following steps:

[0189] Step S902: In response to an input instruction on the operation interface, displaying a structured description text of the target object and an example text corresponding to the structured description text on the operation interface;

[0190] The sample text is used to describe an object of the same type as the target object.

[0191] When a user enters a command or selects a target object on the interface, the system responds by displaying a structured description of the target object and sample text for similar objects. The structured description provides clear information about the target object, while the sample text provides a reference for attributes and content to generate high-quality text.

[0192] Step S904, in response to the processing instruction acting on the operation interface, displaying the target text of the target object on the operation interface;

[0193] Among them, the target text is obtained by using the second text generation model based on the feedback information and multiple initial texts, the feedback information is the feedback information of the multiple initial texts collected by putting the multiple initial texts into the preset scene, the multiple initial texts are input into the first text generation model, and are obtained by using the first text generation model. The multiple initial texts have the same attributes as the sample text, and the target text conforms to the feedback information.

[0194] In response to further user instructions on the user interface, the system displays a target text that has been improved based on user feedback. This target text is generated through the following process: First, the structured description text and sample text are input into the first text generation model, which generates a series of initial texts with the same properties as the sample texts. Subsequently, these initial texts are placed in a preset scenario, and the system collects and analyzes feedback information to determine the user's preference for the text. Finally, based on this feedback information, the second text generation model generates a target text that is closer to the user's preferences. This text is displayed on the user interface for the user to view or use.

[0195] Through the above steps, in response to the input instruction on the operation interface, the structured description text of the target object and the example text corresponding to the structured description text are displayed on the operation interface, wherein the example text is used to describe an object of the same type as the target object; in response to the processing instruction on the operation interface, the target text of the target object is displayed on the operation interface, wherein the target text is obtained by using the second text generation model based on the feedback information and multiple initial texts, the feedback information is the feedback information of the multiple initial texts collected by putting the multiple initial texts into the preset scene, the multiple initial texts are obtained by inputting the structured description text and the example text into the first text generation model, and the multiple initial texts have the same attributes as the example text. The text conforms to the feedback information, and the purpose of improving the text generation effect is achieved; it is easy to notice that a plurality of initial texts with the same attributes as the sample text can be generated by using the first text generation model through the sample text that describes the object of the same type as the target object, and the plurality of initial texts can be put into the preset scene. By analyzing the feedback information, the initial text features with a higher degree of user preference in the preset scene are identified, and the second text generation model is used to adjust its generation strategy, so that the generated target text is more likely to have these features, while reducing or avoiding the initial text features with a lower degree of user preference from appearing in the target text, thereby improving the pertinence and user acceptance of text generation, and thus solving the technical problem of poor text generation effect in related technologies.

[0196] Figure 10 is a flowchart of a text generation method according to an embodiment of the present application, such as Figure 10 As shown, the method may include the following steps:

[0197] Step S1002: acquiring a structured description text of a target object and a sample text corresponding to the structured description text by calling a first interface;

[0198] The first interface includes a first parameter, and the parameter value of the first parameter includes a structured description text and an example text, and the example text is used to describe an object of the same type as the target object.

[0199] The above-mentioned first interface can be an interface for data interaction between the cloud server and the client. The structured description text and the example text corresponding to the structured description text can be passed into the interface function as the first parameter of the interface function to achieve the purpose of uploading the structured description text and the example text corresponding to the structured description text to the cloud server.

[0200] Step S1004: inputting the structured description text and the sample text into a first text generation model, and using the first text generation model to generate a plurality of initial texts corresponding to the target object;

[0201] Among them, the properties of the multiple initial texts are the same as those of the sample texts.

[0202] Step S1006: placing the multiple initial texts into a preset scene and collecting feedback information of the multiple initial texts;

[0203] The feedback information is used to indicate the user's preference for multiple initial texts in a preset scenario.

[0204] Step S1008, generating a target text of the target object using a second text generation model based on the feedback information and the multiple initial texts;

[0205] Among them, the target text conforms to the feedback information.

[0206] Step S1010: Output the target text by calling the second interface.

[0207] The second interface includes a second parameter, and the parameter value of the second parameter includes the target text.

[0208] The above-mentioned second interface can be an interface for data interaction between the cloud server and the client. The cloud server can pass the target text into the interface function as the second parameter of the interface function to achieve the purpose of sending the target text to the client.

[0209] Through the above steps, a structured description text of a target object and an example text corresponding to the structured description text are obtained by calling a first interface, wherein the first interface includes a first parameter, and a parameter value of the first parameter includes a structured description text and an example text, and the example text is used to describe an object of the same type as the target object; the structured description text and the example text are input into a first text generation model, and a plurality of initial texts corresponding to the target object are generated using the first text generation model, wherein the attributes of the plurality of initial texts are the same as those of the example text; the plurality of initial texts are placed in a preset scene, and feedback information of the plurality of initial texts is collected, wherein the feedback information is used to indicate the degree of user preference for the plurality of initial texts in the preset scene; based on the feedback information and the plurality of initial texts, a target text of the target object is generated using a second text generation model, wherein the target text conforms to the feedback information; The target text is output by calling the second interface, wherein the second interface includes a second parameter, and the parameter value of the second parameter includes the target text information, thereby achieving the purpose of improving the text generation effect; it is easy to notice that the sample text that describes the same type of object as the target object can be used to generate multiple initial texts with the same attributes as the sample text using the first text generation model, and the multiple initial texts can be put into the preset scene. By analyzing the feedback information, the initial text features with a higher degree of user preference in the preset scene are identified, and the second text generation model is used to adjust its generation strategy, so that the generated target text is more likely to have these features, while reducing or avoiding the initial text features with a lower degree of user preference from appearing in the target text, thereby improving the pertinence and user acceptance of text generation, and thus solving the technical problem of poor text generation effect in related technologies.

[0210] According to an embodiment of the present application, a text generation device for implementing the above-mentioned text generation method is also provided. Figure 11 is a schematic diagram of a text generation device according to an embodiment of the present application, such as Figure 11 As shown, the device 1100 includes: an acquisition module 1102 , a first generation module 1104 , a collection module 1106 , and a second generation module 1108 .

[0211] Among them, the acquisition module is used to obtain the structured description text of the target object and the example text corresponding to the structured description text, wherein the example text is used to describe an object of the same type as the target object; the first generation module is used to input the structured description text and the example text into the first text generation model, and use the first text generation model to generate multiple initial texts corresponding to the target object, wherein the multiple initial texts have the same attributes as the example text; the collection module is used to put the multiple initial texts into a preset scene, and collect feedback information of the multiple initial texts, wherein the feedback information is used to indicate the user's preference for the multiple initial texts in the preset scene; the second generation module is used to generate the target text of the target object based on the feedback information and the multiple initial texts using the second text generation model, wherein the target text conforms to the feedback information.

[0212] It should be noted that the acquisition module 1102, the first generation module 1104, the acquisition module 1106, and the second generation module 1108 correspond to steps S202 to S208 in the above embodiment. The examples and application scenarios implemented by the four modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the above modules or units can be hardware components or software components stored in a memory and processed by one or more processors. The above modules can also be part of the device and can run in the server 10 provided in the above embodiment.

[0213] In the above embodiment of the present application, the second generation module is also used to select a first text and a second text from multiple initial texts based on feedback information, wherein the user preference level of the first text is greater than the user preference level of the initial texts other than the first text, and the user preference level of the second text is less than the user preference level of the initial texts other than the second text; based on the first text and the second text, the target text is generated using the second text generation model.

[0214] In the above embodiment of the present application, the second generation module is also used to obtain the scene description text of the preset scene and the associated description text of the target object; input the scene description text, the associated description text, the first text and the second text into the second text generation model, and use the second text generation model to generate the target text.

[0215] In the above embodiment of the present application, the second generation module is also used to respond to feedback information of the user interaction amount in a preset scenario, sort multiple initial texts based on the user interaction amount to obtain a sorting result; match the multiple initial texts based on the sorting result to obtain multiple text pairs, wherein the degree of text difference in different text pairs is different; screen the multiple text pairs based on the text difference degree to obtain a target text pair, wherein the text difference degree of the target text pair is greater than the text difference degree of other text pairs in the multiple text pairs except the target text pair, and the target text pair includes a first text and a second text.

[0216] In the above embodiment of the present application, the device is also used to obtain multiple sample texts of a sample object; put the multiple sample texts into multiple sample scenarios, and collect sample feedback information of the multiple sample texts, wherein the sample feedback information is used to represent the user preference level for the multiple sample texts in the multiple sample scenarios; select a first sample text and a second sample text from the multiple sample texts based on the sample feedback information, wherein the user preference level of the first sample text is greater than the user preference level of the sample texts other than the first sample text, and the user preference level of the second sample text is less than the user preference level of the sample texts other than the second sample text; based on the first sample text and the second sample text, generate a target sample text using an initial text generation model; adjust the model parameters of the initial text generation model based on the first sample text, the second sample text and the target sample text to obtain a second text generation model.

[0217] In the above embodiment of the present application, the device is also used to construct a first loss function based on the first sample text and the target sample text, wherein the first loss function is used to increase the similarity between the target sample text and the first sample text; construct a second loss function based on the second sample text and the target sample text, wherein the second loss function is used to reduce the similarity between the target sample text and the second sample text; adjust the model parameters based on the first loss function and the second loss function to obtain a second text generation model.

[0218] In the above embodiment of the present application, the device is also used to obtain sample scene description texts of multiple sample scenes and sample associated description texts of sample objects; input the sample scene description texts, sample associated description texts, first sample texts and second sample texts into the initial text generation model, and use the initial text generation model to generate target sample texts.

[0219] In the above embodiment of the present application, the acquisition module is also used to obtain multimodal data of the target object, wherein the multimodal data is used to describe the target object through different modalities; convert the multimodal data into an initial description text of the target object; perform structured processing on the initial description text to obtain a structured description text; and retrieve sample text from a database based on the type of the target object, wherein the database is used to store sample texts of different types of objects.

[0220] According to an embodiment of the present application, a text generation device for implementing the above-mentioned text generation method is also provided. Figure 12 is a schematic diagram of a text generation device according to an embodiment of the present application, such as Figure 12 As shown, the device 1200 includes: an acquisition module 1202 , a first generation module 1204 , a collection module 1206 , and a second generation module 1208 .

[0221] Among them, the acquisition module is used to obtain the structured description text of the target product and the example title text corresponding to the structured description text, wherein the example title text is used to describe the product of the same type as the target product; the first generation module is used to input the structured description text and the example title text into the first text generation model, and use the first text generation model to generate multiple initial title texts corresponding to the target product, wherein the multiple initial title texts have the same attributes as the example title text; the collection module is used to put the multiple initial title texts into the preset scene, and collect feedback information of the multiple initial title texts, wherein the feedback information is used to indicate the user's preference for the multiple initial title texts in the preset scene; the second generation module is used to generate the target title text of the target product based on the feedback information and the multiple initial title texts using the second text generation model, wherein the target title text conforms to the feedback information.

[0222] It should be noted that the acquisition module 1202, the first generation module 1204, the collection module 1206, and the second generation module 1208 correspond to steps S802 to S808 in the above embodiment. The examples and application scenarios implemented by the four modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the above modules or units can be hardware components or software components stored in a memory and processed by one or more processors. The above modules can also be part of the device and can run in the server 10 provided in the above embodiment.

[0223] According to an embodiment of the present application, a text generation device for implementing the above-mentioned text generation method is also provided. Figure 13 is a schematic diagram of a text generation device according to an embodiment of the present application, such as Figure 13 As shown, the device 1300 includes: a first display module 1302 and a second display module 1304 .

[0224] Among them, the first display module is used to respond to the input instructions acting on the operation interface, and display the structured description text of the target object and the example text corresponding to the structured description text on the operation interface, wherein the example text is used to describe an object of the same type as the target object; the second display module is used to respond to the processing instructions acting on the operation interface, and display the target text of the target object on the operation interface, wherein the target text is obtained by using the second text generation model based on the feedback information and multiple initial texts, the feedback information is the feedback information of the multiple initial texts collected by putting the multiple initial texts into the preset scene, the multiple initial texts are obtained by inputting the structured description text and the example text into the first text generation model, the multiple initial texts have the same attributes as the example text, and the target text conforms to the feedback information.

[0225] It should be noted that the first display module 1302 and the second display module 1304 correspond to steps S902 to S904 in the above embodiment. The examples and application scenarios implemented by the two modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the above modules or units can be hardware components or software components stored in a memory and processed by one or more processors. The above modules can also be part of the device and can run on the server 10 provided in the above embodiment.

[0226] According to an embodiment of the present application, a text generation device for implementing the above-mentioned text generation method is also provided. Figure 14 is a schematic diagram of a text generation device according to an embodiment of the present application, such as Figure 14 As shown, the device 1400 includes: an acquisition module 1402 , a first generation module 1404 , a collection module 1406 , a second generation module 1408 , and an output module 1410 .

[0227] Among them, the acquisition module is used to obtain the structured description text of the target object and the example text corresponding to the structured description text by calling the first interface, wherein the first interface includes a first parameter, the parameter value of the first parameter includes the structured description text and the example text, and the example text is used to describe an object of the same type as the target object; the first generation module is used to input the structured description text and the example text into the first text generation model, and use the first text generation model to generate multiple initial texts corresponding to the target object, wherein the multiple initial texts have the same attributes as the example text; the collection module is used to put the multiple initial texts into a preset scene, and collect feedback information of the multiple initial texts, wherein the feedback information is used to indicate the user's preference for the multiple initial texts in the preset scene; the second generation module is used to generate the target text of the target object based on the feedback information and the multiple initial texts using the second text generation model, wherein the target text conforms to the feedback information; the output module is used to output the target text by calling the second interface, wherein the second interface includes a second parameter, and the parameter value of the second parameter includes the target text.

[0228] It should be noted that the acquisition module 1402, the first generation module 1404, the acquisition module 1406, the second generation module 1408, and the output module 1410 correspond to steps S1002 to S1010 in the above embodiment. The examples and application scenarios implemented by the five modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above embodiment. It should be noted that the above modules or units can be hardware components or software components stored in a memory and processed by one or more processors. The above modules can also be part of the device and can run in the server 10 provided in the above embodiment.

[0229] It should be noted that the preferred implementation scheme involved in the above embodiments of the present application is the same as the scheme provided in the above embodiments, as well as the application scenario and implementation process, but is not limited to the scheme provided in the above embodiments.

[0230] An embodiment of the present application may provide a computing device. Figure 15 This is a structural block diagram of a computing device according to an embodiment of the present application. Figure 15 As shown, the computing device 100 may include: one or more ( Figure 15 (only one is shown) processor 102, memory 104, memory controller, and peripheral interfaces.

[0231] The above-mentioned computing device can be understood as an integrated intelligent terminal, including but not limited to a server, a desktop computer, a PC (Personal Computer), a model all-in-one machine, etc., and the computing device can be pre-installed with the model described in the above-mentioned embodiment of this application.

[0232] Specifically, the computing device can pre-install multiple types of models, including but not limited to models in the fields of natural language processing, visual processing, speech processing, code processing, and multimodal task processing, thereby providing a diverse selection of models. In different product forms, the computing device can support one or more model usage methods, including but not limited to model training, model calling, model fine-tuning, model deployment, model reasoning and application, etc. In some product forms, the computing device also supports model management, including but not limited to multi-type model management (supporting the management of multiple types of models such as discriminants and genesis), model version control (supporting the control of different model versions), model evaluation (based on model evaluation tools to evaluate the performance and effectiveness of the model), etc. In other product forms, the computing device can also create applications based on the model and provide API calling capabilities. The model can be called into the created application through the API interface, and application management tools are provided to achieve management and monitoring of the application.

[0233] Furthermore, the computing device can also include data management (supporting the creation and management of model tuning data sets), a training center (providing rich training resources to help users learn and master AI technology), and basic management and control capabilities (providing enterprise-level basic management and control capabilities to ensure the security and efficient operation of the system). Through the above functions, a comprehensive, integrated AI development, training, deployment and application device is provided.

[0234] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the methods in the above embodiments. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely located relative to the processor, and these remote memories may be connected to terminal A via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0235] The processor may call the executable program stored in the memory through the transmission device to execute the method described in any one of the above embodiments.

[0236] An embodiment of the present application may provide an electronic device. Figure 16 This is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 16As shown, the electronic device may include: an input / output device 112 ; a memory 114 ; and a processor 116 , wherein the processor 116 is connected to the input / output device 112 and the memory 114 via a bus 118 .

[0237] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the methods in the above embodiments. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely located relative to the processor, and these remote memories may be connected to terminal A via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0238] The processor may call the executable program stored in the memory through the transmission device to execute the method described in any one of the above embodiments.

[0239] Those skilled in the art will understand that Figure 16 The structure shown is for illustrative purposes only. The computing device may also be a smartphone (e.g., an Android phone, an iOS phone, etc.), a tablet computer, a PDA, or a terminal device such as a mobile internet device (MID) or PAD. This figure does not limit the structure of the computing device described above. For example, computing device 100 may include more or fewer components (e.g., a network interface, a display device, etc.) than shown in this figure, or may have a configuration different from that shown in this figure.

[0240] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0241] The embodiment of the present application further provides a computer-readable storage medium. Optionally, in this embodiment, the computer-readable storage medium can be used to store the program code executed by the method provided in the above embodiment.

[0242] Optionally, in this embodiment, the above storage medium may be located in a computing device.

[0243] Optionally, in this embodiment, the computer-readable storage medium is configured to store an executable program, and when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the method described in any one of the above embodiments.

[0244] The embodiment of the present application further provides a computer program product. Optionally, in this embodiment, the computer program product may include a computer program, and when the computer program is executed by a processor, the method provided in the embodiment is implemented.

[0245] The embodiments of the present application further provide a computer program product. Optionally, the computer program product may include a non-volatile computer-readable storage medium, which may be used to store a computer program that, when executed by a processor, implements the method provided in the embodiments above.

[0246] The embodiment of the present application further provides a computer program. Optionally, in this embodiment, when the computer program is executed by a processor, the method provided in the above embodiment is implemented.

[0247] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0248] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0249] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0250] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0251] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program code.

[0252] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A text generation method, characterized in that: include: Acquire a structured description text of a target object and an example text corresponding to the structured description text, wherein the example text is used to describe an object of the same type as the target object; Inputting the structured description text and the sample text into a first text generation model, and using the first text generation model to generate a plurality of initial texts corresponding to the target object, wherein the plurality of initial texts have the same attributes as the sample text; placing the plurality of initial texts into different preset scenarios and collecting feedback information of the plurality of initial texts, wherein the feedback information is used to indicate a user's preference for the plurality of initial texts in the preset scenarios; selecting a first text and a second text from the plurality of initial texts based on the feedback information, wherein a user preference level for the first text is greater than a user preference level for initial texts other than the first text, and a user preference level for the second text is less than a user preference level for initial texts other than the second text; Based on the scene description text of the preset scene, the associated description text of the target object, the first text and the second text, a target text of the target object is generated using a second text generation model, wherein the target text conforms to the feedback information, and the associated description text is used to represent other types of description text associated with the target object.

2. The method according to claim 1, characterized in that Selecting a first text and a second text from the plurality of initial texts based on the feedback information includes: In response to the feedback information being the amount of user interaction in the preset scenario, sorting the plurality of initial texts based on the amount of user interaction to obtain a sorting result; Matching the multiple initial texts based on the sorting results to obtain multiple text pairs, wherein the text differences in different text pairs are different; The multiple text pairs are screened based on the text difference degree to obtain a target text pair, wherein the text difference degree of the target text pair is greater than the text difference degrees of other text pairs in the multiple text pairs except the target text pair, and the target text pair includes the first text and the second text.

3. The method according to claim 1, characterized in that The method further comprises: Get multiple sample texts of the sample object; placing the plurality of sample texts into a plurality of sample scenarios, and collecting sample feedback information of the plurality of sample texts, wherein the sample feedback information is used to indicate user preference for the plurality of sample texts in the plurality of sample scenarios; selecting a first sample text and a second sample text from the plurality of sample texts based on the sample feedback information, wherein the user preference level for the first sample text is greater than the user preference levels for sample texts other than the first sample text, and the user preference level for the second sample text is less than the user preference levels for sample texts other than the second sample text; Based on the first sample text and the second sample text, generating a target sample text using an initial text generation model; The model parameters of the initial text generation model are adjusted based on the first sample text, the second sample text and the target sample text to obtain the second text generation model.

4. The method according to claim 3, characterized in that Adjusting the model parameters of the initial text generation model based on the first sample text, the second sample text, and the target sample text to obtain the second text generation model includes: Constructing a first loss function based on the first sample text and the target sample text, wherein the first loss function is used to increase the similarity between the target sample text and the first sample text; Constructing a second loss function based on the second sample text and the target sample text, wherein the second loss function is used to reduce the similarity between the target sample text and the second sample text; The model parameters are adjusted based on the first loss function and the second loss function to obtain the second text generation model.

5. The method according to claim 3, characterized in that Generating a target sample text using an initial text generation model based on the first sample text and the second sample text includes: Obtaining sample scene description texts of the plurality of sample scenes and sample associated description texts of the sample objects; The sample scene description text, the sample association description text, the first sample text, and the second sample text are input into the initial text generation model, and the target sample text is generated using the initial text generation model.

6. The method according to claim 1, characterized in that Obtaining a structured description text of a target object and an example text corresponding to the structured description text, including: Acquiring multimodal data of the target object, wherein the multimodal data is used to describe the target object in different modalities; Converting the multimodal data into an initial description text of the target object; Performing structural processing on the initial description text to obtain the structured description text; Based on the type of the target object, the sample text is retrieved from a database, wherein the database is used to store sample texts of objects of different types.

7. A text generation method, characterized in that: include: Obtaining a structured description text of a target product and an example title text corresponding to the structured description text, wherein the example title text is used to describe a product of the same type as the target product; Inputting the structured description text and the example title text into a first text generation model, and using the first text generation model to generate a plurality of initial title texts corresponding to the target product, wherein the plurality of initial title texts have the same attributes as the example title text; placing the plurality of initial title texts in different preset scenarios, and collecting feedback information on the plurality of initial title texts, wherein the feedback information is used to indicate a user's preference for the plurality of initial title texts in the preset scenarios; selecting a first text and a second text from the plurality of initial title texts based on the feedback information, wherein a user preference level for the first text is greater than a user preference level for initial title texts other than the first text, and a user preference level for the second text is less than a user preference level for initial title texts other than the second text; Based on the scene description text of the preset scene, the associated description text of the target product, the first text and the second text, a target title text of the target product is generated using a second text generation model, wherein the target title text conforms to the feedback information, and the associated description text is used to represent other types of description text associated with the target product.

8. A text generation method, characterized in that: include: In response to an input instruction applied to an operation interface, displaying a structured description text of a target object and an example text corresponding to the structured description text on the operation interface, wherein the example text is used to describe an object of the same type as the target object; In response to a processing instruction acting on the operation interface, a target text of the target object is displayed on the operation interface, wherein the target text is a scene description text based on a preset scene, an associated description text of the target object, a first text and a second text obtained using a second text generation model, the first text and the second text are selected from a plurality of initial texts based on feedback information, the plurality of initial texts are obtained by inputting the structured description text and the sample text into the first text generation model using the first text generation model, the feedback information is feedback information of the plurality of initial texts collected by placing the plurality of initial texts into different preset scenes, the plurality of initial texts have the same attributes as the sample text, the target text conforms to the feedback information, the associated description text is used to represent other types of description texts associated with the target object, the user preference level of the first text is greater than the user preference level of the initial texts other than the first text, and the user preference level of the second text is less than the user preference level of the initial texts other than the second text.

9. A text generation method, characterized in that: include: Acquire a structured description text of a target object and an example text corresponding to the structured description text by calling a first interface, wherein the first interface includes a first parameter, a parameter value of the first parameter includes the structured description text and the example text, and the example text is used to describe an object of the same type as the target object; Inputting the structured description text and the sample text into a first text generation model, and using the first text generation model to generate a plurality of initial texts corresponding to the target object, wherein the plurality of initial texts have the same attributes as the sample text; placing the plurality of initial texts into different preset scenarios and collecting feedback information of the plurality of initial texts, wherein the feedback information is used to indicate a user's preference for the plurality of initial texts in the preset scenarios; selecting a first text and a second text from the plurality of initial texts based on the feedback information, wherein a user preference level for the first text is greater than a user preference level for initial texts other than the first text, and a user preference level for the second text is less than a user preference level for initial texts other than the second text; generating a target text for the target object using a second text generation model based on the scene description text of the preset scene, the associated description text of the target object, the first text, and the second text, wherein the target text conforms to the feedback information, and the associated description text is used to represent other types of description text associated with the target object; The target text is output by calling a second interface, wherein the second interface includes a second parameter, and a parameter value of the second parameter includes the target text.

10. An electronic device, characterized in that: include: a memory storing an executable program; A processor is connected to the memory via a bus, and is configured to run the program, wherein the program executes the method according to any one of claims 1 to 9 when running.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored executable program, wherein when the executable program is run, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 9.

12. A computer program product, characterized in that The invention comprises a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 9.

Images