Product review generation method, device, equipment and storage medium

Abstract

The present application discloses a method, apparatus, device and storage medium for generating product reviews, which relates to the field of artificial intelligence technology, including: obtaining initial review information corresponding to a target product, and using a preset prompt word generation template to generate prompt words corresponding to the target product based on the initial review information, the product information of the target product and the user information of the target user; determining the target features corresponding to the initial review information based on a preset multi-head attention model, and generating a large model based on the prompt words, the target features and the preset reviews to generate a pending review corresponding to the target product; obtaining the evaluation information generated by the target user based on the pending review, and determining the target review corresponding to the target product based on the evaluation information. In this way, the present application can simplify the steps of product reviews, reduce the workload and difficulty of writing product reviews, help users effectively complete product reviews, and improve the quality of product reviews.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, device, equipment and storage medium for generating product reviews. Background Art

[0002] Existing shopping website user review data is generally self-written, and the quality of these reviews varies widely. Most users find it cumbersome to fill out numerous reviews, so they fill them out casually. This results in a high number of invalid reviews, which in turn fail to reflect users' true feelings and significantly impact subsequent product marketing. Current user product review analysis systems mostly analyze the review data after users have completed their comments, a form of post-hoc analysis. These systems typically classify product review data, prioritizing high-quality reviews for user reference, thereby influencing their purchasing decisions. Existing review quality analysis techniques typically rely on existing review data and then use intelligent analysis techniques to assess the effectiveness of the reviews.

[0003] In summary, how to simplify the steps of product reviews and thus improve the quality of existing product reviews is a technical problem that needs to be solved urgently. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a product review generation method, device, equipment and storage medium that can simplify the steps of product review generation and thus improve the quality of existing product reviews. The specific solution is as follows:

[0005] In a first aspect, the present application provides a method for generating product reviews, comprising:

[0006] Obtaining initial review information corresponding to a target product, and using a preset prompt word generation template to generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user;

[0007] Determine the target feature corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt word, the target feature and the preset review to generate a review to be confirmed corresponding to the target product;

[0008] The evaluation information generated by the target user based on the comment to be confirmed is obtained, and a target comment corresponding to the target product is determined based on the evaluation information.

[0009] Optionally, the generating of the prompt word corresponding to the target product according to the initial review information, the product information of the target product, and the user information of the target user using a preset prompt word generation template includes:

[0010] Generate the initial review information corresponding to the target product based on a preset review template or a review writing instruction of a target user;

[0011] Determining a first target text corresponding to the initial review information and a second target text corresponding to the user information of the target user, and determining a preset marketing attribute corresponding to the target product;

[0012] The preset prompt word generation template is used to generate the prompt word corresponding to the target product according to the first target text, the second target text, the product information of the target product, and the preset marketing attributes.

[0013] Optionally, before determining the target feature corresponding to the initial review information based on the preset multi-head attention model, the method further includes:

[0014] Determine a first target word vector corresponding to the first target text, the second target text, and the product information based on a preset word embedding technology;

[0015] Features of the first target word vector are extracted according to a preset converter encoder model to obtain first initial features corresponding to the initial review information, the user information, and the product information.

[0016] Optionally, before determining the target feature corresponding to the initial review information based on the preset multi-head attention model, the method further includes:

[0017] Determining a target video and a target picture that may be included in the initial comment information, and extracting a target key frame in the target video based on a preset frame difference extraction method;

[0018] Extracting first image features corresponding to the target keyframe and the target image based on a preset convolutional neural network model, and fusing the first image features according to a preset model construction method to obtain corresponding second image features;

[0019] A preset principal component analysis method is used to perform corresponding dimensionality reduction processing on the second image feature, and a third image feature corresponding to the second image feature after dimensionality reduction is determined according to a preset visual model.

[0020] Optionally, before determining the target feature corresponding to the initial review information based on the preset multi-head attention model, the method further includes:

[0021] Determine a third target text in the target video according to a preset speech recognition method, and determine text features corresponding to the third target text according to a preset text convolutional neural network model;

[0022] Determining a second initial feature corresponding to the initial comment information based on a preset cross-attention mechanism, the text feature, and the third image feature;

[0023] Accordingly, determining the target features corresponding to the initial review information based on the preset multi-head attention model includes:

[0024] Determining corresponding features to be fused based on the preset multi-head attention model, the preset normalized exponential function, the first initial features, and the second initial features;

[0025] The feature to be fused and the first initial feature are fused according to the preset model construction method to obtain the target feature corresponding to the initial comment information.

[0026] Optionally, generating a large model based on the prompt word, the target feature, and preset comments to generate a to-be-confirmed comment corresponding to the target product includes:

[0027] Based on the preset word embedding technology, the second target word vector corresponding to the prompt word is determined, and the second target word vector and the target feature are input into the preset comment generation model to generate the to-be-confirmed comment corresponding to the target product.

[0028] Optionally, determining a target review corresponding to the target product based on the evaluation information includes:

[0029] Directly determining the review to be confirmed as the target review corresponding to the target product based on the evaluation information;

[0030] Or, based on the evaluation information, adjust the random number factor in the preset comment generation model, and jump to the step of generating the large model based on the prompt word, the target feature and the preset comment, and generate the pending comment corresponding to the target product to obtain the target comment corresponding to the target product.

[0031] In a second aspect, the present application provides a product review generation device, comprising:

[0032] a prompt word generation module, configured to obtain initial review information corresponding to a target product, and generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user using a preset prompt word generation template;

[0033] A pending review generation module is configured to determine the target feature corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt word, the target feature, and the preset review to generate a pending review corresponding to the target product;

[0034] The target review determination module is used to obtain the evaluation information generated by the target user based on the review to be confirmed, and determine the target review corresponding to the target product based on the evaluation information.

[0035] In a third aspect, the present application provides an electronic device, comprising:

[0036] Memory, used to store computer programs;

[0037] A processor is used to execute the computer program to implement the aforementioned product review generation method.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned product review generation method is implemented.

[0039] In this application, the initial review information corresponding to the target product is first obtained, and a preset prompt word generation template is used to generate prompt words corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user; then, the target features corresponding to the initial review information are determined based on the preset multi-head attention model, and a large model is generated based on the prompt words, the target features, and the preset comments to generate a pending review corresponding to the target product; finally, the evaluation information generated by the target user based on the pending review is obtained, and the target review corresponding to the target product is determined based on the evaluation information. As can be seen from the above, this application can perform real-time analysis and judgment on the initial review information, generate corresponding prompt words in combination with product information and user information, and use the preset comment generation large model to assist in writing the user's product review based on the prompt words and initial review information, thereby simplifying the steps of product review, reducing the workload and difficulty of writing product review, leveling the cultural level and age differences between users, helping users to effectively complete product reviews, and improving the quality of product reviews. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0041] Figure 1 A schematic diagram of the system framework applicable to a product review generation solution provided in this application;

[0042] Figure 2 A flowchart of a product review generation method provided in this application;

[0043] Figure 3 A schematic diagram of a specific personalized review template provided for this application;

[0044] Figure 4 A specific multimodal hybrid feature extraction flow chart provided for this application;

[0045] Figure 5 A specific automatic comment completion diagram provided for this application;

[0046] Figure 6 A flowchart of a specific product review generation method provided in this application;

[0047] Figure 7 A specific shopping review large language model training flow chart provided for this application;

[0048] Figure 8 A flowchart of a specific product review generation method provided in this application;

[0049] Figure 9 A schematic diagram of the structure of a product review generation device provided in this application;

[0050] Figure 10 This is a structural diagram of an electronic device provided in this application. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0052] The shopping evaluation data of users of existing shopping websites are generally written by themselves, and the quality of the reviews varies. Most users find it troublesome to fill in a lot of reviews, so they fill them in casually, resulting in a lot of invalid reviews. A large number of product reviews cannot reflect the real feelings of users, which has a relatively large impact on the subsequent marketing of the products. Most of the current user product review analysis systems conduct judgment and analysis on the evaluation data after the user comments, which is a post-analysis. Basically, they classify and identify the product review data, and put high-quality reviews in the front row for user reference, so as to achieve the purpose of influencing the user's shopping desire. Existing review quality analysis technologies are mostly based on existing review data, and then use intelligent analysis technology to judge the effect of the reviews. To this end, the present application provides a product review generation solution that can simplify the steps of product reviews, thereby improving the quality of existing product reviews.

[0053] The system framework used in the product review generation solution of the present invention can be found in Figure 1 As shown. A large language model for shopping reviews is constructed by generating vertical domains of shopping mall shopping reviews with marketing attributes, which is mainly used for the automatic generation of product reviews of mall users on the C-end. And when designing the large model, factors such as marketing type, marketing level, and user emotions are taken into consideration to affect the product reviews that are finally generated. In addition, the present invention selects the LLaVa model (LLaVa, Large Language and Vision Assistant) as the basic multimodal large language model, improves the input of the model on the basis of the LLaVa model, and performs end-to-end fine-tuning training on the model, so that the LLaVa model has the ability to review products on the basis of general knowledge.

[0054] In the system framework of the present invention, the "multimodal hybrid feature extraction and encoding" module is mainly responsible for feature extraction of multimodal data. The parameters of the feature extraction encoder are frozen. When training the large shopping review language model, only the Projection layer (i.e., the mapping learning weight layer) is trained. It is mainly used to align the multimodal features with the word vectors of the pre-trained large shopping review language model.

[0055] The product review collection template and product reviews correspond to two different ways of entering initial review information. When a user directly leaves a review, their product review is directly fed into the "Shopping Review Large Language Model Prompt Word Instruction" module. Alternatively, the product review collection template collects user information and then feeds it into the "Shopping Review Large Language Model Prompt Word Instruction" module. The system also uses marketing attributes to set the marketing type and level of the review, which are then fed into the "Shopping Review Large Language Model Prompt Word Instruction" module to construct the final Prompt instruction for the shopping review large language model.

[0056] The "Shopping Review Large Language Model Prompt Word Instructions" module is a prompt word engineering tool that guides and controls the behavior of the shopping review large language model, enabling it to generate more accurate and targeted product reviews. Prompt word engineering primarily guides the shopping review large language model's product review behavior by defining tasks and context settings, while also controlling the marketing characteristics and quality of the content output through product marketing attributes. The "Shopping Review Large Language Model Prompt Word Instructions" module constructs product review instructions corresponding to the target product. Then, through word vectorization, the input prompt instructions are converted into corresponding word vectors. These are then input into the shopping review large language model along with the feature vectors generated by the "Multimodal Hybrid Feature Extraction and Encoding" model. The fine-tuned shopping review large language model then generates the final product review. Finally, users can provide feedback on the product review through conversation, enabling the shopping review large language model to generate more appropriate results.

[0057] See also Figure 2 As shown, an embodiment of the present invention discloses a method for generating product reviews, which may include:

[0058] Step S11: Acquire initial review information corresponding to a target product, and use a preset prompt word generation template to generate a prompt word corresponding to the target product according to the initial review information, product information of the target product, and user information of the target user.

[0059] In the present embodiment, the above-mentioned use of a preset prompt word generation template to generate prompt words corresponding to the target product according to the initial review information, the product information of the target product, and the user information of the target user may include: generating the initial review information corresponding to the target product based on a preset review template or a review writing instruction of a target user; respectively determining the first target text corresponding to the initial review information and the second target text corresponding to the user information of the target user, and determining the preset marketing attributes corresponding to the target product; using the preset prompt word generation template to generate the prompt words corresponding to the target product according to the first target text, the second target text, the product information of the target product, and the preset marketing attributes. Specifically, the initial review information can be determined through the product review written directly by the user, or when the quality of the user review is too poor or the user actively initiates the generation of a review, the comment auxiliary generation function can be selected, and then a personalized review template for the product review can be popped up to quickly collect user usage experience information and obtain the corresponding initial review information, wherein the personalized review template for the product review can be found in Figure 3As shown, users can summarize their user experience by giving ratings on user experience, design, quality, and selecting subjective evaluation words. Afterwards, the text portion of the initial review information and user information is extracted to obtain the corresponding first target text and second target text, and the marketing type, marketing level, etc. of the target review are set through marketing attributes, wherein user information includes but is not limited to: age group, gender, address location, tags, previous reviews, etc. Then, a preset prompt word generation template is used to generate the corresponding Prompt instruction, i.e., prompt word, based on the first target text, second target text, product information, and marketing attributes. Among them, the preset prompt word generation template can be specifically referred to as shown below:

[0060] Instructions: Please generate a personalized user review of a product based on the following input data;

[0061] Context:

[0062] The user's basic information tags (such as male, single, young people from small towns, car owners, etc.);

[0063] Basic information about the purchased goods;

[0064] Product experience (quality satisfaction (neutral, unsatisfactory), appearance liking (neutral, unsatisfactory), enjoyment of the experience, etc.);

[0065] Examples of good reviews for this product;

[0066] Marketing attributes (marketing type, marketing level, etc.);

[0067] Input data (this field is optional if the user chooses to use a personalized review template to collect their feedback, or if the user enters partial product review text and the system is waiting to generate additional data, the user-entered review data will be used here);

[0068] Output instructions: Specify the output format of product reviews.

[0069] Step S12: determine the target features corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt words, the target features and the preset reviews to generate the pending reviews corresponding to the target product.

[0070] See also Figure 4As shown, in this embodiment, a multimodal hybrid feature extraction method is designed to integrate features from text, images, and videos, thereby providing more basic feature information for the shopping review language model. Both fine-tuning and inference of the shopping review language model require the use of multimodal hybrid feature extraction techniques. A Transformer Encoder model (i.e., a transformer encoder model) can be used for text feature extraction. After fine-tuning the text using word embedding technology, a text feature vector is extracted. Features from the review images and review videos are then extracted as image feature vectors. The text feature vectors and image feature vectors are then fused together using a fully connected layer, thereby combining the features of the two modalities into a unified feature vector. Specifically, the data requiring feature extraction is first obtained, including product information, initial review information, review images, review videos, and user information. Except for product information, the remaining data is optional and feature extraction is performed when provided by the user.

[0071] In a specific embodiment, in order to extract text features corresponding to product information, initial review information, and user information, before determining the target features corresponding to the initial review information based on the preset multi-head attention model, the method may also include: determining the first target word vector corresponding to the first target text, the second target text, and the product information based on the preset word embedding technology; extracting the features of the first target word vector according to the preset transformer encoder model to obtain the first initial features corresponding to the initial review information, the user information, and the product information. Specifically, the text data of the obtained product information, initial review information, and user information are converted into corresponding vector representations through Word Embedding (i.e., word embedding) technology, and then the word vector is input into the Transformer Encoder model to extract the corresponding text features, thereby obtaining the first initial feature. The generated text feature vector A of the first initial feature can be used with F 1 Indicates that F 1 is a 1024-dimensional feature vector. In this embodiment, the feature extraction of user information and product information will be performed after obtaining user authorization to protect user privacy and improve the review generation efficiency of the shopping review large language model.

[0072] In a specific embodiment, in order to extract image features corresponding to comment pictures and comment videos, before determining the target features corresponding to the initial comment information based on the preset multi-head attention model, it may also include: determining the target video and target picture that may be contained in the initial comment information, and extracting the target key frame in the target video based on the preset frame difference extraction method; extracting the first image features corresponding to the target key frame and the target picture based on the preset convolutional neural network model, and fusing the first image features according to the preset model construction method to obtain the corresponding second image features; performing corresponding dimensionality reduction processing on the second image features using the preset principal component analysis method, and determining the third image features corresponding to the second image features after dimensionality reduction according to the preset visual model. It can be understood that comment pictures and comment videos are optional data, and feature extraction is performed when provided by the user. For the target pictures of the comments obtained, assuming that the user has uploaded k pictures, where k>=0, the system will perform feature extraction on the k pictures respectively. Specifically, the ResNext-50 model can be used to extract image features. The features generated for each picture can be expressed using P i Indicates that the image feature F can be obtained by directly concatenating the combination of image features through the Concat operation 2 , then the features of the k-frame image can be expressed as F 2 =(P 1 , P 2 ,…,P k ), F 2 is a feature vector of dimensions T*2048*k, where T is the size of the image, which is generally 64*64. For the target video of the reviews, since the product review videos are generally short and there is a lot of redundancy in the video information, it is designed to extract only the key frames of the video, and extract a total of j frames of images as key frames. Specifically, the frame difference extraction method can be used to extract the key frames of the video, and the target frame f i With the previous frame f i-1 Perform image difference. When the difference is greater than the threshold T, it is determined that the target frame has a large image change, and the target frame is considered to be a key frame, which can be expressed as: Then the extracted video key frame images are subjected to image feature extraction. The features of the video image can be expressed as F 3 =(P 1 , P 2 ,…,P j ). Afterwards, for the image feature vector F 2 , video image feature vector F 3Since the number of comment images k and the number of video key frames j are uncertain, we can use PCA dimensionality reduction (PCA, i.e., Principal Component Analysis) to extract the core feature vector. Then, we concatenate the first image feature of the target image and the second image feature of the target video to obtain F 4 , that is, F 4 =(F 2 , F 3 ). Then, 4 Perform PCA dimensionality reduction to obtain the main features G of the target image and target video of the comment 1 , G 1 The feature vector is T*1024 dimensional, and the image features of k+j frames of the target image and target video are de-redundant and fixed to the feature vector of T*1024 dimension. Finally, the Transformer-vit model (i.e., VisionTransformer) is used to transform the feature G 1 Perform fusion and feature extraction to determine the semantic coding expression of the target image and target video to facilitate subsequent feature fusion.

[0073] In a specific embodiment, in order to extract the speech and text features corresponding to the comment video, before determining the target features corresponding to the initial comment information based on the preset multi-head attention model, it can also include: determining the third target text in the target video according to the preset speech recognition method, and determining the text features corresponding to the third target text according to the preset text convolutional neural network model; determining the second initial features corresponding to the initial comment information based on the preset cross-attention mechanism, the text features and the third image features. Specifically, the speech of the target video can be converted into text information through speech recognition to obtain the third target text. If there is no audio information in the target video, this step can be ignored. The third target text is input into the TextCNN model (Text Convolutional Neural Network, i.e., text convolutional neural network model), and the corresponding text feature information is extracted as T 1 , T 1 It is a 768-dimensional feature vector, and then the feature fusion is performed through the CrossAttenion (i.e., cross attention mechanism) structure to obtain the fused image feature vector B, which is a 1*1024-dimensional feature vector.

[0074] In this embodiment, the determination of the target features corresponding to the initial comment information based on the preset multi-head attention model may include: determining the corresponding features to be fused based on the preset multi-head attention model, the preset normalized exponential function, the first initial features, and the second initial features; fusing the features to be fused and the first initial features according to the preset model construction method to obtain the target features corresponding to the initial comment information. Specifically, a multi-head attention model is constructed to fuse the text feature vector A of the first initial feature with the image feature vector B of the second initial feature. The value V in the attention mechanism is obtained by transforming the text feature vector A, and the query Q and the key value K are obtained by transforming the image feature vector B.

[0075] ;

[0076] ;

[0077] ;

[0078] The formula represents the construction of the feature vector of the i-th attention mechanism, where W v i , W q i , W k i Respectively represent the weight transfer matrix of the nonlinear transformation features of the hidden layer feature vector, + represents the addition of each column of the matrix to the corresponding vector. v i , b q i , b k i Represents the bias vector of the model. In this embodiment, the attention mechanism uses 6 heads, and the formula for fusing text features with image features is as follows:

[0079] ;

[0080] ;

[0081] Q in the formula i , K i 、V i Represents the query, keyword, and value in each spatial attention mechanism in the multi-head attention mechanism. Through the cross-attention mechanism, the text features and image features are deeply integrated to obtain the feature vector H i , and then the feature vectors of several heads are combined with the weight transfer matrix W of the nonlinear transformation feature i The matrix is ​​multiplied and the heads of each attention are accumulated. In order to avoid the model explosion, the output data can be normalized to between 0 and 1 through the Softmax operation to obtain the feature vector G of the fused feature to be fused. 2 , G 2is a 1*1024-dimensional feature vector. In order to prevent the forgetting of text features, the feature vector G of the feature to be fused can be 2 The text feature vector F with the first initial feature 1 A Concat operation is performed to obtain the feature vector C of the final fused target feature, where C is a 1*2048-dimensional feature vector. This shows that this embodiment uses multimodal fusion technology to enable the shopping review large language model to understand user comments, photos, and video information. By leveraging multimodal fusion features, it can more naturally understand user product reviews, thereby extracting more effective auxiliary features to help the shopping review large language model generate user product reviews. The resulting product reviews are of higher quality and more consistent with the user's actual usage experience, while also possessing some auxiliary marketing attributes.

[0082] In this embodiment, the generation of a large model based on the prompt word, the target feature and the preset comment generation to generate the pending comment corresponding to the target product may include: determining the second target word vector corresponding to the prompt word based on the preset word embedding technology, and inputting the second target word vector and the target feature into the preset comment generation large model to generate the pending comment corresponding to the target product. Specifically, the input Prompt instruction can first be converted into a corresponding word vector through word vectorization, and then input into the shopping review large language model together with the feature vector of the target feature, and the pending comment of the target product is generated by the fine-tuned shopping review large language model, wherein the feature vector of the target feature is used as the guiding feature vector of the shopping review large language model for inference to guide the shopping review large language model to generate product reviews that meet the basic information of the user. See Figure 5 As shown, in one specific embodiment, when a user enters the product review interface and starts to review a product, the system can automatically complete the subsequent review text for the user, and the completed review text to be confirmed is gray. The user can use the left and right keys on the keyboard to select the content to be accepted, and can choose to accept all or partially accept. If it is partially accepted, the system will continue to generate a new one for the accepted part and let the user make a choice. In another specific embodiment, the user's usage experience information is quickly collected through the personalized review template of the product review to obtain the corresponding initial review information, and then the prompt words of the large language model are constructed based on the collected initial review information, and then sent to the shopping review large language model to generate the review to be confirmed for the target product.

[0083] Step S13: obtaining evaluation information generated by the target user based on the comment to be confirmed, and determining a target comment corresponding to the target product based on the evaluation information.

[0084] In this embodiment, the user can judge the pending comments generated by the shopping review large language model through dialogue, so that the large language model can give a more appropriate generation result. If the pending comments are consistent with the user's feelings, the user can accept the generated product review. If the generated pending comments do not meet the user's subjective usage experience of the product, the user can input text evaluation of the generated pending comments, or directly select the review experience provided by the system, such as consistent with subjective feelings, worse usage experience than the pending comments, better usage experience than the pending comments, etc.

[0085] As can be seen from the above, in this embodiment, the initial review information corresponding to the target product is first obtained, and a preset prompt word generation template is used to generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user; then, the target feature corresponding to the initial review information is determined based on the preset multi-head attention model, and a large model is generated based on the prompt word, the target feature, and the preset review to generate a pending review corresponding to the target product; finally, the evaluation information generated by the target user based on the pending review is obtained, and the target review corresponding to the target product is determined based on the evaluation information. As can be seen from the above, this embodiment can perform real-time analysis and judgment on the initial review information, generate corresponding prompt words in combination with product information and user information, and use the preset review generation large model to assist in writing the user's product review based on the prompt word and initial review information, thereby simplifying the steps of product review, reducing the workload and difficulty of writing product review, leveling the cultural level and age differences between users, helping users to effectively complete product reviews, and improving the quality of product reviews.

[0086] See also Figure 6 As shown, in order to obtain product reviews that satisfy users, the embodiment of the present invention further discloses a product review generation method, which may include:

[0087] Step S21: Acquire initial review information corresponding to a target product, and use a preset prompt word generation template to generate a prompt word corresponding to the target product according to the initial review information, product information of the target product, and user information of the target user.

[0088] Step S22: determine the target features corresponding to the initial review information based on the preset multi-head attention model, and generate a large model based on the prompt words, the target features and the preset reviews to generate the pending reviews corresponding to the target product.

[0089] See also Figure 7As shown in the figure, the training process of the shopping review large language model in this embodiment is as follows: First, nearly one million product review data from existing shopping malls is collected. This data is cleaned and graded according to marketing effectiveness to construct a basic data set. The model is pre-trained using the basic data set. Through self-supervised training on this basic data, the model learns the semantics of basic knowledge text and the logical relationships between sentences.

[0090] During the model's intensive training phase, human judgement feedback is used to evaluate the quality of generated text, improving the model's ability to generate reviews that are close to user experience across a wide range of product reviews and guiding model optimization. Specifically, the model generates each generated result, which is then passed on to human judgement, generating rewards and penalties, thereby enabling the model to learn the output habits expected by humans.

[0091] Finally, the model is fine-tuned and trained using labeled product review information data, and the model parameters are fine-tuned using sensitive data to optimize the model's performance in the user product review generation task, thereby obtaining the final shopping review language model.

[0092] Step S23: obtaining evaluation information generated by the target user based on the comment to be confirmed, and directly determining the comment to be confirmed as the target review corresponding to the target product based on the evaluation information.

[0093] In one specific implementation, users directly enter product reviews, and the system analyzes them in real time. If a review is incomplete, the system automatically completes it for the user. If the user accepts the review, the target review for the target product can be directly determined based on the pending reviews generated by the shopping review large language model.

[0094] In another specific implementation, if the quality of user reviews is too poor or the user actively initiates the generation of reviews, a personalized template for product reviews will pop up to guide the user to quickly collect usage experience information, and then generate new product reviews based on user experience. If the user adopts the pending review generated by the shopping review large language model, the pending review can be directly used to overwrite the original review.

[0095] Step S24: obtain the evaluation information generated by the target user based on the comment to be confirmed, adjust the random number factor in the preset comment generation model based on the evaluation information, and jump to the step of generating a large model based on the prompt word, the target feature and the preset comment to generate the comment to be confirmed corresponding to the target product, so as to obtain the target comment corresponding to the target product.

[0096] In this embodiment, if the user rejects the pending comment, the system will automatically adjust the random number factor of the shopping review language model based on the user's evaluation information, and regenerate product review information that is close to the user's actual usage experience, and so on until the user is satisfied and accepts it, and then the user will publish the target review.

[0097] For a more specific processing procedure of the above step S21, reference may be made to the corresponding contents disclosed in the above embodiments, which will not be described again here.

[0098] As can be seen above, this embodiment first trains a shopping review large language model using existing product review data. After using the shopping review large language model to generate a pending review for the target product, the user evaluates the pending review. If the user accepts the pending review, the target review can be directly obtained based on the pending review. If the user rejects the review, the shopping review large language model is adjusted and the product review is regenerated until the user is satisfied and accepts it. In this way, by learning from high-quality user product reviews, the shopping review large language model is equipped with the ability to generate high-quality product reviews. This does not rely on manual input or cultural level of the user, and only requires simple selection and judgment to complete the product usage evaluation. Moreover, this embodiment can generate product reviews based on user evaluations that are close to the user's actual usage experience.

[0099] See also Figure 8 As shown, in a specific implementation, the product review generation process may be specifically as follows:

[0100] a) When a user starts a review, they can either write a product review directly or select the review assistance function. When a user writes a product review directly, the system quickly analyzes the review and provides real-time completion.

[0101] b) When review quality is poor, enable the assisted review generation feature. Or if the user prefers not to fill out the form themselves, they can proactively enable this feature. This feature will generate different feedback collection templates based on the product category the user purchased, with specific decisions made based on the product. Users simply need to check the questions in the template and briefly describe their experience. Once the template is completed, the system will generate a product review based on the user's information.

[0102] c) The information collected in the above two steps will serve as the basic input data for the shopping review large language model. At the same time, the user's personalized information, product information of the purchased goods, and user-uploaded photos and videos will be collected and integrated into the feature to construct the prompt word instructions of the large model. Finally, the information is sent to the shopping review large language model for analysis through the cross-attention mechanism to generate new user product review information.

[0103] d) Users will evaluate the generated product reviews. If the reviews are consistent with the user's feelings, the user can accept the generated product review information. If the generated product review information does not meet the user's subjective feelings about the product, the user can enter a text evaluation of the generated product review information or directly select the review feeling provided by the system.

[0104] e) If a user rejects a product review, the system automatically adjusts the random number factor based on the user's comments, regenerating a product review that reflects the user's actual experience. This process continues until the user is satisfied and accepts the review, at which point the user can publish the review. In practical testing, users typically reach satisfaction and acceptance after one or two revisions.

[0105] Accordingly, see Figure 9 As shown, the embodiment of the present application also provides a product review generation device, which may include:

[0106] The prompt word generation module 11 is used to obtain the initial review information corresponding to the target product, and use a preset prompt word generation template to generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user;

[0107] The pending review generation module 12 is configured to determine the target feature corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt word, the target feature and the preset review to generate a pending review corresponding to the target product;

[0108] The target review determination module 13 is configured to obtain the evaluation information generated by the target user based on the review to be confirmed, and determine the target review corresponding to the target product based on the evaluation information.

[0109] As can be seen from the above, in this application, the initial review information corresponding to the target product is first obtained, and a preset prompt word generation template is used to generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product and the user information of the target user; then, the target feature corresponding to the initial review information is determined based on the preset multi-head attention model, and a large model is generated based on the prompt word, the target feature and the preset review to generate a pending review corresponding to the target product; finally, the evaluation information generated by the target user based on the pending review is obtained, and the target review corresponding to the target product is determined based on the evaluation information. As can be seen from the above, this application can perform real-time analysis and judgment on the initial review information, generate corresponding prompt words in combination with product information and user information, and use the preset review generation large model to assist in writing the user's product review based on the prompt word and initial review information, thereby simplifying the steps of product review, reducing the workload and difficulty of writing product review, leveling the cultural level and age differences between users, helping users to effectively complete product reviews, and improving the quality of product reviews.

[0110] In some specific implementations, the prompt word generation module 11 may include:

[0111] An initial review information generating unit, configured to generate the initial review information corresponding to the target product based on a preset review template or a review writing instruction of a target user;

[0112] a preset marketing attribute determination unit, configured to respectively determine a first target text corresponding to the initial review information and a second target text corresponding to the user information of the target user, and determine a preset marketing attribute corresponding to the target product;

[0113] The prompt word generating unit is configured to generate the prompt word corresponding to the target product according to the first target text, the second target text, the product information of the target product, and the preset marketing attributes by using the preset prompt word generating template.

[0114] In some specific implementations, the product review generating device may further include:

[0115] A first target word vector determination module, configured to determine first target word vectors corresponding to the first target text, the second target text, and the product information based on a preset word embedding technology;

[0116] A first initial feature determination module is used to extract features of the first target word vector according to a preset converter encoder model to obtain first initial features corresponding to the initial review information, the user information, and the product information.

[0117] In some specific implementations, the product review generating device may further include:

[0118] A target key frame extraction module is used to determine the target video and target image that may be included in the initial comment information, and extract the target key frame in the target video based on a preset frame difference extraction method;

[0119] a second image feature determination module, configured to extract first image features corresponding to the target keyframe and the target image based on a preset convolutional neural network model, and fuse the first image features according to a preset model construction method to obtain corresponding second image features;

[0120] The third image feature determination module is used to perform corresponding dimensionality reduction processing on the second image feature using a preset principal component analysis method, and determine the third image feature corresponding to the second image feature after dimensionality reduction according to a preset visual model.

[0121] In some specific implementations, the product review generating device may further include:

[0122] a text feature determination module, configured to determine a third target text in the target video according to a preset speech recognition method, and determine text features corresponding to the third target text according to a preset text convolutional neural network model;

[0123] A second initial feature determination module is configured to determine a second initial feature corresponding to the initial comment information based on a preset cross-attention mechanism, the text feature, and the third image feature;

[0124] Accordingly, the to-be-confirmed comment generating module 12 may include:

[0125] a feature-to-be-fused determining unit, configured to determine corresponding features to be fused based on the preset multi-head attention model, the preset normalized exponential function, the first initial features, and the second initial features;

[0126] A target feature determination unit is configured to fuse the feature to be fused and the first initial feature according to the preset model construction method to obtain the target feature corresponding to the initial comment information.

[0127] In some specific implementations, the to-be-confirmed comment generating module 12 may include:

[0128] The pending comment generation unit is used to determine the second target word vector corresponding to the prompt word based on the preset word embedding technology, and input the second target word vector and the target feature into the preset comment generation model to generate the pending comment corresponding to the target product.

[0129] In some specific implementations, the target review determination module 13 may include:

[0130] a first target review determining unit, configured to directly determine the to-be-confirmed review as the target review corresponding to the target product based on the evaluation information;

[0131] The second target comment determination unit is used to adjust the random number factor in the preset comment generation model based on the evaluation information, and jump to the step of generating the large model according to the prompt word, the target feature and the preset comment, and generating the to-be-confirmed comment corresponding to the target product, so as to obtain the target comment corresponding to the target product.

[0132] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 10 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram cannot be considered as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the product review generation method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0133] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0134] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0135] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the product review generation method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 may further include a computer program capable of performing other specific tasks.

[0136] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned method for generating product reviews. The specific steps of this method can be referred to the corresponding contents disclosed in the aforementioned embodiments and will not be repeated here.

[0137] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0138] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0139] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0140] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0141] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for generating product reviews, characterized in that: include: Obtaining initial review information corresponding to a target product, and using a preset prompt word generation template to generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user; Determine the target feature corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt word, the target feature and the preset review to generate a review to be confirmed corresponding to the target product; Obtaining evaluation information generated by the target user based on the review to be confirmed, and determining a target review corresponding to the target product based on the evaluation information; The step of using a preset prompt word generation template to generate a prompt word corresponding to the target product according to the initial review information, the product information of the target product, and the user information of the target user includes: Generate the initial review information corresponding to the target product based on a preset review template or a review writing instruction of a target user; Determining a first target text corresponding to the initial review information and a second target text corresponding to the user information of the target user, and determining a preset marketing attribute corresponding to the target product; The preset prompt word generation template is used to generate the prompt word corresponding to the target product according to the first target text, the second target text, the product information of the target product, and the preset marketing attributes.

2. The method for generating product reviews according to claim 1, wherein: Before determining the target features corresponding to the initial review information based on the preset multi-head attention model, the method further includes: Determine a first target word vector corresponding to the first target text, the second target text, and the product information based on a preset word embedding technology; Features of the first target word vector are extracted according to a preset converter encoder model to obtain first initial features corresponding to the initial review information, the user information, and the product information.

3. The method for generating product reviews according to claim 2, wherein: Before determining the target features corresponding to the initial review information based on the preset multi-head attention model, the method further includes: Determining a target video and a target picture included in the initial comment information, and extracting a target key frame in the target video based on a preset frame difference extraction method; Extracting first image features corresponding to the target keyframe and the target image based on a preset convolutional neural network model, and fusing the first image features according to a preset model construction method to obtain corresponding second image features; A preset principal component analysis method is used to perform corresponding dimensionality reduction processing on the second image feature, and a third image feature corresponding to the second image feature after dimensionality reduction is determined according to a preset visual model.

4. The method for generating product reviews according to claim 3, wherein: Before determining the target features corresponding to the initial review information based on the preset multi-head attention model, the method further includes: Determine a third target text in the target video according to a preset speech recognition method, and determine text features corresponding to the third target text according to a preset text convolutional neural network model; Determining a second initial feature corresponding to the initial comment information based on a preset cross-attention mechanism, the text feature, and the third image feature; Accordingly, determining the target features corresponding to the initial review information based on the preset multi-head attention model includes: Determining corresponding features to be fused based on the preset multi-head attention model, the preset normalized exponential function, the first initial features, and the second initial features; The feature to be fused and the first initial feature are fused according to the preset model construction method to obtain the target feature corresponding to the initial comment information.

5. The method for generating product reviews according to claim 1, wherein: The step of generating a large model based on the prompt word, the target feature, and the preset comments, and generating a comment to be confirmed corresponding to the target product, includes: Based on the preset word embedding technology, the second target word vector corresponding to the prompt word is determined, and the second target word vector and the target feature are input into the preset comment generation model to generate the to-be-confirmed comment corresponding to the target product.

6. The method for generating product reviews according to any one of claims 1 to 5, wherein: The determining a target review corresponding to the target product based on the evaluation information includes: Directly determining the review to be confirmed as the target review corresponding to the target product based on the evaluation information; Or, based on the evaluation information, adjust the random number factor in the preset comment generation model, and jump to the step of generating the large model based on the prompt word, the target feature and the preset comment, and generate the pending comment corresponding to the target product to obtain the target comment corresponding to the target product.

7. A product review generating device, characterized in that: include: A prompt word generation module is used to obtain initial review information corresponding to a target product, and generate a prompt word corresponding to the target product based on the initial review information, the product information of the target product, and the user information of the target user using a preset prompt word generation template; A pending review generation module is configured to determine the target feature corresponding to the initial review information based on a preset multi-head attention model, and generate a large model based on the prompt word, the target feature, and the preset review to generate a pending review corresponding to the target product; a target review determination module, configured to obtain evaluation information generated by the target user based on the review to be confirmed, and determine a target review corresponding to the target product based on the evaluation information; The prompt word generation module includes: An initial review information generating unit, configured to generate the initial review information corresponding to the target product based on a preset review template or a review writing instruction of a target user; a preset marketing attribute determination unit, configured to respectively determine a first target text corresponding to the initial review information and a second target text corresponding to the user information of the target user, and determine a preset marketing attribute corresponding to the target product; The prompt word generating unit is configured to generate the prompt word corresponding to the target product according to the first target text, the second target text, the product information of the target product, and the preset marketing attributes by using the preset prompt word generating template.

8. An electronic device, characterized in that: The electronic device includes a processor and a memory; wherein the memory is used to store a computer program, and the computer program is loaded and executed by the processor to implement the product review generation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that Used to store a computer program, which, when executed by a processor, implements the product review generating method according to any one of claims 1 to 6.

Images