Data recommendation method and device, computer equipment and storage medium

Abstract

The embodiment of the invention belongs to the field of financial science and technology, and relates to a data recommendation method and device, computer equipment and a storage medium, and the method comprises the following steps: extracting commodity metadata and user historical behavior data from a commodity transaction log; performing vectorization representation on the commodity metadata and the user historical behavior data to obtain a commodity feature vector and a user feature vector; performing similarity matching according to the commodity feature vector and the user feature vector to obtain a recommended commodity set; performing feature extraction on the recommended commodity set, and inputting the feature to a pre-trained commodity evaluation model to obtain a recommended commodity score; generating a recommended commodity list according to the scores of the recommended commodities, and performing clustering and grouping to obtain recommended commodity groups; and adjusting the recommended commodity group according to the real-time behavior data of the user and carrying out data recommendation. According to the method, accurate and personalized data recommendation can be effectively carried out on the user, so that the conversion rate of the user is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, specifically to the field of financial technology, and in particular to a data recommendation method, device, computer equipment and storage medium. Background Art

[0002] In the current financial business platform ecosystem, product display is a key bridge connecting merchants and consumers, and its effect directly affects user participation and conversion rate. In the traditional model, platforms often adopt a unified layout and strategy to pile up a wide range of products on the homepage or specific category pages, hoping to attract users' attention and facilitate transactions. However, this "one-size-fits-all" display method ignores the differences among individual users, including their purchasing needs, personal preferences, consumption habits, and even psychological expectations.

[0003] With the rapid development of technologies such as big data and artificial intelligence, financial business platforms have gradually realized the importance of personalized recommendation systems. Traditional product display methods lack precision, which often leads to a large number of users quickly losing interest in browsing or even becoming bored when faced with product information they are not interested in. This not only reduces user stay time and interaction frequency, but also seriously restricts users from converting to actual purchasing behavior. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a data recommendation method, apparatus, computer device and storage medium to solve the problem of being unable to make accurate and personalized data recommendations to users.

[0005] In order to solve the above technical problems, the present application embodiment provides a data recommendation method, which adopts the following technical solution:

[0006] Obtaining commodity transaction logs, and extracting commodity metadata and user historical behavior data from the commodity transaction logs;

[0007] Vectorizing the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector;

[0008] Performing similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set;

[0009] Extracting features from the recommended product set to obtain recommended product attribute features, and inputting the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score;

[0010] Generating a recommended product list according to the recommended product scores, and clustering and grouping the recommended product list to obtain a recommended product group;

[0011] Real-time user behavior data is obtained, the recommended product group is adjusted according to the real-time user behavior data to obtain a valid recommended product group, and data recommendation is performed according to the valid recommended product group.

[0012] Furthermore, the step of obtaining the commodity transaction log and extracting commodity metadata and user historical behavior data from the commodity transaction log specifically includes:

[0013] Get the transaction log extraction identifier;

[0014] Extracting the commodity transaction log from the database according to the transaction log extraction identifier;

[0015] The commodity transaction log is parsed to obtain the commodity metadata and the user historical behavior data.

[0016] Furthermore, the step of vectorizing the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector specifically includes:

[0017] Extracting features from the product metadata and the user historical behavior data to obtain product interaction features and user historical behavior features;

[0018] Performing a first dimensionality reduction process on the commodity interaction feature to obtain the commodity feature vector;

[0019] A second dimensionality reduction process is performed on the user historical behavior features to obtain the user feature vector.

[0020] Furthermore, the step of performing similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set specifically includes:

[0021] Performing cosine similarity calculation on the product feature vector based on a collaborative filtering algorithm to obtain product similarity, and generating a first recommended product based on the product similarity;

[0022] Performing cosine similarity calculation on the user feature vector based on a collaborative filtering algorithm to obtain user similarity, and generating a second recommended product according to the user similarity;

[0023] The data of the first recommended product and the second recommended product are integrated to obtain the recommended product set.

[0024] Furthermore, the step of extracting features from the recommended product set to obtain recommended product attribute features, and inputting the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score specifically includes:

[0025] Preprocessing the recommended product set to obtain a standard recommended product set;

[0026] Acquire commodity attribute information, and perform feature extraction on the standard recommended commodity set according to the commodity attribute information to obtain attribute features of the recommended commodity;

[0027] A model extraction identifier is obtained, the product evaluation model is extracted from a database according to the model extraction identifier, and the attribute characteristics of the recommended product are input into the product evaluation model to obtain the recommended product score.

[0028] Furthermore, the step of generating a recommended product list according to the recommended product scores, and clustering and grouping the recommended product list to obtain a recommended product group specifically includes:

[0029] Sort the products according to the recommended product scores to obtain a recommended product ranking table;

[0030] Obtaining a recommendation score threshold, and filtering the recommended product ranking table according to the recommendation score threshold to obtain the recommended product list;

[0031] Clustering and grouping the recommended product list according to a clustering algorithm to obtain a recommended product cluster;

[0032] Obtaining category information of the recommended product cluster, and obtaining a category sorting strategy according to the category information;

[0033] The recommended product clusters are sorted and screened according to the category sorting strategy to obtain the recommended product group.

[0034] Furthermore, the step of obtaining the user's real-time behavior data and adjusting the recommended product group according to the user's real-time behavior data to obtain a valid recommended product group specifically includes:

[0035] Acquire a real-time behavior data extraction identifier, and extract the user's real-time behavior data from a database according to the real-time behavior data extraction identifier;

[0036] Performing behavior analysis on the real-time behavior data of the user to obtain user behavior information;

[0037] The recommended product group is matched with products, prioritized, and optimized according to the user behavior information to obtain the effective recommended product group.

[0038] In order to solve the above technical problems, the embodiment of the present application further provides a data recommendation device, which adopts the following technical solution:

[0039] A data acquisition module, used to acquire commodity transaction logs, and extract commodity metadata and user historical behavior data from the commodity transaction logs;

[0040] A data conversion module, used to vectorize the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector;

[0041] A data matching module, used to perform similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set;

[0042] A data evaluation module is used to extract features from the recommended product set to obtain recommended product attribute features, and input the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score;

[0043] A data classification module, used to generate a recommended product list according to the recommended product scores, and cluster the recommended product list to obtain a recommended product group;

[0044] The data recommendation module is used to obtain real-time user behavior data, adjust the recommended product group according to the real-time user behavior data to obtain a valid recommended product group, and perform data recommendation based on the valid recommended product group.

[0045] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0046] A computer device comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of any of the above data recommendation methods when executing the computer-readable instructions.

[0047] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0048] A computer-readable storage medium having computer-readable instructions stored thereon, wherein the computer-readable instructions, when executed by a processor, implement the steps of any of the above data recommendation methods.

[0049] Compared with the prior art, the embodiment of the present application has the following beneficial effects: the embodiment obtains commodity transaction logs, extracts commodity metadata and user historical behavior data from the commodity transaction logs; vectorizes the commodity metadata and the user historical behavior data to obtain commodity feature vectors and user feature vectors; performs similarity matching based on the commodity feature vectors and the user feature vectors to obtain a recommended commodity set; extracts features from the recommended commodity set to obtain recommended commodity attribute features, and inputs the recommended commodity attribute features into a pre-trained commodity evaluation model to obtain a recommended commodity score; generates a recommended commodity list based on the recommended commodity score, and clusters and groups the recommended commodity list to obtain a recommended commodity group; obtains user real-time behavior data, adjusts the recommended commodity group based on the user real-time behavior data to obtain an effective recommended commodity group, and performs data recommendation based on the effective recommended commodity group. Thus, accurate and personalized data recommendations are effectively made to users to improve the user conversion rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0052] Figure 2 A flowchart of an embodiment of a data recommendation method according to the present application;

[0053] Figure 3 yes Figure 2 A flowchart of a specific implementation of step S10;

[0054] Figure 4 yes Figure 2 A flowchart of a specific implementation of step S20;

[0055] Figure 5 yes Figure 2 A flowchart of a specific implementation of step S30;

[0056] Figure 6 yes Figure 2 A flowchart of a specific implementation of step S40;

[0057] Figure 7 yes Figure 2 A flowchart of a specific implementation of step S50;

[0058] Figure 8 yes Figure 2 A flowchart of a specific implementation of step S60;

[0059] Fig. 9 is a structural schematic diagram of an embodiment of a data recommendation device according to the present application;

[0060] Fig.10 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0061] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0062] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor does it refer to non-related or alternative embodiments that are mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0063] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0064] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0065] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0066] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, desktop computers, etc.

[0067] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .

[0068] It should be noted that the data recommendation method provided in the embodiment of the present application is generally executed by a server, and accordingly, the data recommendation device is generally arranged in the server.

[0069] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0070] Continue to refer Figure 2 , shows a flow chart of an embodiment of a method for data recommendation according to the present application. The data recommendation method comprises the following steps:

[0071] Step S10, obtaining a commodity transaction log, and extracting commodity metadata and user historical behavior data from the commodity transaction log;

[0072] In this embodiment, the commodity transaction log includes basic transaction information: transaction time: the specific date and time when the transaction occurs; information of both parties to the transaction: including the name, ID or account information of the seller and the buyer; commodity information: the name, specification, model, quantity, unit price, etc. of the commodity; transaction amount: the total amount of the transaction, which may include the total price of the commodity, taxes, freight, etc.; payment method: the payment method used, such as Alipay, WeChat payment, bank card payment, etc. Transaction process information: order number: the order number that uniquely identifies each transaction; order status: records the various stages of the order, such as pending payment, paid, pending shipment, shipped, completed, cancelled, etc.; transaction interaction information: communication records: communication records between the buyer and seller, such as chat content, negotiation results, etc.; complaints and feedback: complaints, evaluations or feedback information on the transaction. Financial and settlement information: income and expenditure: detailed records of income and expenditure in the transaction; settlement time: settlement time and method of transaction funds. Security and risk information: abnormal transaction warning: abnormal transaction behavior identified by the system or manually; risk assessment: assessment results of transaction risks.

[0073] Step S20, vectorizing the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector;

[0074] In this embodiment, commodity metadata refers to data describing commodities, including basic commodity information: such as commodity name, commodity number, brand, manufacturer, place of origin, etc.; commodity description information: including detailed description of commodities, specifications, functional characteristics, instructions for use, etc.; price and promotion information: selling price, original price, discount information, promotion activities, etc. of commodities; inventory and logistics information: inventory quantity, shipping place, expected arrival time, etc. of commodities; classification and label: classification, label, keyword, etc. of commodities. User historical behavior data refers to the behavior data generated by users in the process of accessing and using the system platform, including browsing behavior: the list of commodities browsed by users, page dwelling time, browsing path, etc.; search behavior: search keywords entered by users, search time, click status of search results, etc.; collection and purchase behavior: commodities collected by users, commodities added to shopping carts, etc.; purchase behavior: user purchase records, purchase time, purchase amount, purchase quantity, etc.; evaluation and feedback behavior: user evaluation, rating, comment content, feedback, etc. of commodities. The product feature vector and the user feature vector are feature vectors corresponding to the product metadata and the user historical behavior data, and can be obtained by performing feature extraction and dimensionality reduction processing on the product metadata and the user historical behavior data.

[0075] Step S30, performing similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set;

[0076] In this embodiment, the recommended product set is a data set generated based on similarity matching between the product feature vector and the user feature vector. The recommended product set includes product data obtained by matching the product feature vector and the user feature vector.

[0077] Step S40, extracting features from the recommended product set to obtain recommended product attribute features, and inputting the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score;

[0078] In this embodiment, the pre-trained product evaluation model adopts a logistic regression model, and the training steps of the logistic regression model include: obtaining sample product data; performing data cleaning, noise removal, and outlier processing on the sample product data to obtain standard sample product data; performing feature extraction on the standard sample product data to obtain sample product features; setting the layer structure of the logistic regression model and initializing the model parameters; training the logistic regression model using the sample product features as a training data set, calculating the loss function, and optimizing the logistic regression model using the gradient descent method to obtain a product rating model.

[0079] Step S50, generating a recommended product list according to the recommended product scores, and clustering the recommended product list to obtain a recommended product group;

[0080] In this embodiment, the recommended product score is the evaluation score corresponding to the product in the recommended product set, the recommended product list is the list information obtained by sorting and filtering according to the recommended product score, and clustering grouping refers to clustering the recommended product list according to a preset clustering algorithm. In this embodiment, the preset clustering algorithm can adopt the K-Means algorithm. The K-Means algorithm is a widely used clustering algorithm. It divides the data set into K clusters in an iterative manner so that the data points in each cluster are as similar as possible, and the data points between different clusters are as different as possible. The K-Means algorithm can be used to cluster the products according to their characteristics (such as price, sales volume, user evaluation, etc.), so as to identify product groups with similar attributes.

[0081] Step S60, obtaining user real-time behavior data, adjusting the recommended product group according to the user real-time behavior data to obtain a valid recommended product group, and performing data recommendation according to the valid recommended product group.

[0082] In this embodiment, the effective recommended product group refers to the product information output by the system to the display page for data recommendation display. After obtaining the effective recommended product group, the system sends the effective recommended product group to the corresponding user's browsing display page for data display, thereby completing the data recommendation for the user.

[0083] This embodiment obtains commodity transaction logs, extracts commodity metadata and user historical behavior data from the commodity transaction logs; vectorizes the commodity metadata and the user historical behavior data to obtain commodity feature vectors and user feature vectors; performs similarity matching based on the commodity feature vectors and the user feature vectors to obtain a recommended commodity set; extracts features from the recommended commodity set to obtain recommended commodity attribute features, and inputs the recommended commodity attribute features into a pre-trained commodity evaluation model to obtain recommended commodity scores; generates a recommended commodity list based on the recommended commodity scores, and clusters and groups the recommended commodity list to obtain a recommended commodity group; obtains user real-time behavior data, adjusts the recommended commodity group based on the user real-time behavior data to obtain an effective recommended commodity group, and performs data recommendation based on the effective recommended commodity group. This effectively performs accurate and personalized data recommendations to users to improve user conversion rates.

[0084] Continue to refer Figure 3 In some optional implementations of this embodiment, step S10 includes the following steps:

[0085] Step S101, obtaining a transaction log extraction identifier;

[0086] In this embodiment, the transaction log extraction identifier is a unique identifier of the corresponding commodity transaction log, and the transaction log extraction identifier may be a log ID or a UUID (universally unique identifier).

[0087] Step S102, extracting the commodity transaction log from the database according to the transaction log extraction identifier;

[0088] In this embodiment, a matching query is performed in the database through the transaction log extraction identifier, and the matched information is extracted to obtain the commodity transaction log. In this embodiment, the commodity transaction log is generated by the system collecting and recording various information of the user's interaction with the commodity, and is stored in the database and updated in real time.

[0089] Step S103, parsing the commodity transaction log to obtain the commodity metadata and the user historical behavior data;

[0090] In this embodiment, the commodity transaction log can be parsed and information extracted according to the first query information corresponding to the commodity metadata and the second query information corresponding to the user historical behavior data, so as to effectively obtain the commodity metadata and the user historical behavior data.

[0091] This embodiment obtains a transaction log extraction identifier; extracts the commodity transaction log from a database according to the transaction log extraction identifier; and parses the commodity transaction log, thereby effectively obtaining commodity metadata and user historical behavior data in the commodity transaction log to facilitate subsequent vectorized representation processing.

[0092] Continue to refer Figure 4 In some optional implementations of this embodiment, step S20 includes the following steps:

[0093] Step S201, extracting features from the product metadata and the user historical behavior data to obtain product interaction features and user historical behavior features;

[0094] In this embodiment, before extracting features from the product metadata and user historical behavior data, it is also necessary to perform data cleaning, noise removal, and outlier processing on the product metadata and user historical behavior data to obtain valid product metadata and user historical behavior data. Product interaction features include product click features, product browsing features, product purchase features, product collection features, and product evaluation features. User historical behavior features include user historical browsing features, user historical purchase features, user historical search features, user historical stay features, and user historical evaluation features.

[0095] Step S202, performing a first dimensionality reduction process on the commodity interaction feature to obtain the commodity feature vector;

[0096] In this embodiment, the first dimensionality reduction process may adopt principal component analysis (PCA), and the process of principal component analysis (PCA) includes data standardization: norm =(X-μ) / σ, where X norm is the data matrix after data standardization, X is the original matrix of product interaction features, μ is the mean vector of each feature, σ is the standard deviation vector of each feature (or the diagonal element of the covariance matrix), and the product interaction features are standardized so that the mean of each feature is 0 and the variance is 1; calculate the covariance matrix: S = 1 / (n-1)·X T norm ·X norm , where X T norm For X norm The matrix transpose of n is the number of samples. The covariance matrix measures the correlation between the features in the data set. For n samples, each sample has p features. The covariance matrix is ​​a p×p matrix, where each element C ij Represents the covariance between the i-th feature and the j-th feature. Calculate the eigenvalues ​​and eigenvectors of the covariance matrix: Sv i =λ i·v i , where λ i is the ith eigenvalue, v i is the corresponding eigenvector. The eigenvalue indicates the degree of data change in the direction of the corresponding eigenvector (i.e., variance), while the eigenvector gives the direction of these changes. Select the principal component: Select the eigenvectors corresponding to the first k largest eigenvalues ​​according to the size of the eigenvalue. These eigenvectors are the directions of the principal components. Usually, the number of principal components k selected will be smaller than the number of original features p to achieve the purpose of dimensionality reduction. Data projection: X pca =X norm ·[v1,v2,…,v k ], where X pca It is a data matrix after dimensionality reduction, with k columns. The original data is projected onto the selected principal component to obtain the product feature vector after dimensionality reduction.

[0097] Step S203: performing a second dimensionality reduction process on the user historical behavior features to obtain the user feature vector.

[0098] In this embodiment, the second dimensionality reduction process may also use principal component analysis (PCA), which is used to effectively reduce the dimensionality of the user's historical behavior characteristics to obtain a user feature vector.

[0099] This embodiment extracts features from the product metadata and the user historical behavior data to obtain product interaction features and user historical behavior features; performs a first dimensionality reduction process on the product interaction features to obtain the product feature vector; and performs a second dimensionality reduction process on the user historical behavior features to obtain the user feature vector. Thus, feature extraction and dimensionality reduction are performed effectively based on the product metadata and the user historical behavior data to effectively obtain the product feature vector and the user feature vector.

[0100] Continue to refer Figure 5 In some optional implementations of this embodiment, step S30 includes the following steps:

[0101] Step S301, performing cosine similarity calculation on the product feature vector based on a collaborative filtering algorithm to obtain product similarity, and generating a first recommended product according to the product similarity;

[0102] In this embodiment, the collaborative filtering algorithm is a recommendation algorithm that completely relies on the behavioral relationship between users and products. Its basic principle is to analyze the user's behavior and preferences, find the similarity between users or the similarity between products, and then generate recommended products based on the similarity. In this embodiment, cosine similarity is used to calculate the similarity between different product feature vectors. Cosine similarity measures the degree of proximity of two product feature vectors in direction without considering their size, thereby helping us find which products are similar in user interaction patterns. After obtaining the product similarity, based on the product similarity, other products similar to the products they liked before are recommended to the current user, and the first recommended product is generated by selecting the N products that are most similar to the target product.

[0103] Step S302, performing cosine similarity calculation on the user feature vector based on a collaborative filtering algorithm to obtain user similarity, and generating a second recommended product according to the user similarity;

[0104] In this embodiment, the collaborative filtering algorithm is a recommendation algorithm that completely relies on the behavioral relationship between users and products. Its basic principle is to find the similarity between users or products by analyzing the user's behavior and preferences, and then generate recommended products based on the similarity. In this embodiment, after obtaining the user similarity, based on the user similarity, a part of users who are most similar to the target user is selected as a similar user set; the ratings of similar users are used in combination with the similarity weights to predict the ratings of the target user for the unrated products; and based on the predicted ratings, a second recommended product is generated for the target user.

[0105] Step S303: integrating data of the first recommended product and the second recommended product to obtain the recommended product set.

[0106] In this embodiment, a recommended product set is generated by merging the first recommended product and the second recommended product.

[0107] This embodiment calculates the cosine similarity of the product feature vector based on a collaborative filtering algorithm to obtain product similarity, and generates a first recommended product based on the product similarity; calculates the cosine similarity of the user feature vector based on a collaborative filtering algorithm to obtain user similarity, and generates a second recommended product based on the user similarity; integrates the data of the first recommended product and the second recommended product, thereby effectively obtaining a recommended product set that includes both products similar to the target product and products favored by users similar to the target user, so as to facilitate subsequent feature extraction processing.

[0108] Continue to refer Figure 6 In some optional implementations of this embodiment, step S40 includes the following steps:

[0109] Step S401, preprocessing the recommended product set to obtain a standard recommended product set;

[0110] In this embodiment, preprocessing includes deduplication: checking whether there are duplicate products in the recommended product set, and if so, deduplication is performed; data cleaning: removing products that are missing key information (such as price, title, image link, etc.), or filling in this information (for example, using default values ​​or estimating based on the average value of other similar products); format standardization: ensuring that the format of all product data is consistent, such as price format, date format, etc.; screening: further screening the product set according to business needs or user preferences, such as filtering by price range, brand, category, etc.

[0111] Step S402, obtaining product attribute information, and performing feature extraction on the standard recommended product set according to the product attribute information to obtain the recommended product attribute features;

[0112] In this embodiment, the product attribute information includes a variety of information such as the product's text description, title, price, brand, user reviews, sales volume, and ratings. The recommended product attribute characteristics include the recommended product price, recommended product brand awareness, recommended product user rating average, recommended product praise rate, etc.

[0113] Step S403, obtaining a model extraction identifier, extracting the product evaluation model from a database according to the model extraction identifier, and inputting the recommended product attribute characteristics into the product evaluation model to obtain the recommended product score.

[0114] In this embodiment, the model extraction identifier is the unique identification information corresponding to the product evaluation model. A matching query is performed in the database according to the model extraction identifier to obtain the product evaluation model. In this embodiment, the product evaluation model adopts a pre-trained deep learning model. The deep learning model can adopt a BERT model. The BERT model is a pre-trained language representation model based on the Transformer structure. It extracts key information from text data such as product descriptions and user reviews, and generates corresponding representation vectors, which are then used to predict the evaluation score of the product and obtain the recommended product score.

[0115] This embodiment obtains a standard recommended product set by preprocessing the recommended product set; obtains product attribute information, and extracts features of the standard recommended product set based on the product attribute information to obtain the recommended product attribute features; obtains a model extraction identifier, extracts the product evaluation model from a database based on the model extraction identifier, and inputs the recommended product attribute features into the product evaluation model, thereby obtaining a recommended product score that effectively evaluates the recommended product set, so as to facilitate the subsequent generation of a recommended product list.

[0116] Continue to refer Figure 7 In some optional implementations of this embodiment, step S50 includes the following steps:

[0117] Step S501, sorting the products according to the recommended product scores to obtain a recommended product sorting table;

[0118] In this embodiment, the recommended product set is sorted from high to low according to the scores of the recommended products, thereby obtaining a recommended product sorting table.

[0119] Step S502, obtaining a recommendation score threshold, and filtering the recommended product ranking table according to the recommendation score threshold to obtain the recommended product list;

[0120] In this embodiment, the recommendation score threshold is a pre-set screening threshold. The recommendation score threshold can be set according to the average value of the recommended product scores corresponding to the recommended product ranking table. By comparing the recommended product scores corresponding to the recommended products in the recommended product ranking table with the recommendation score threshold, the recommended products whose recommended product scores are greater than or equal to the recommendation score threshold are obtained, thereby obtaining a recommended product list.

[0121] Step S503, clustering and grouping the recommended product list according to a clustering algorithm to obtain recommended product clusters;

[0122] In this embodiment, the clustering algorithm adopts the K-Means algorithm. By taking the products in the recommended product list as the input of the K-Means algorithm and executing the K-Means algorithm, the clustering results output by the K-Means algorithm are used as the recommended product clusters, thereby achieving effective clustering grouping of the recommended product list.

[0123] Step S504, obtaining category information of the recommended product cluster, and obtaining a category sorting strategy according to the category information;

[0124] In this embodiment, the category information is the category label corresponding to the recommended product cluster, which can be defined by product attributes. The category sorting strategy is a sorting and screening method set according to business needs or user preferences. The category sorting strategy can sort according to popularity, user interest matching, product diversity, etc., and screen according to the preset number of product displays.

[0125] Step S505: sort and filter the recommended product clusters according to the category sorting strategy to obtain the recommended product group.

[0126] In this embodiment, the recommended commodity clusters are sorted and screened by using a category sorting strategy, and the sorted and screened recommended commodity clusters are used as components of the recommended commodity group, thereby finally obtaining the recommended commodity group.

[0127] This embodiment obtains a recommended product ranking table by sorting products according to the recommended product scores; obtains a recommendation score threshold, and filters the recommended product ranking table according to the recommendation score threshold to obtain the recommended product list; clusters and groups the recommended product list according to a clustering algorithm to obtain recommended product clusters; obtains category information of the recommended product clusters, and obtains a category sorting strategy according to the category information; sorts and filters the recommended product clusters according to the category sorting strategy, thereby effectively obtaining high-quality, highly relevant and diversified recommended product groups, so as to facilitate subsequent adjustment and processing based on the recommended product groups.

[0128] Continue to refer Figure 8 In some optional implementations of this embodiment, step S60 includes the following steps:

[0129] Step S601, obtaining a real-time behavior data extraction identifier, and extracting the user's real-time behavior data from a database according to the real-time behavior data extraction identifier;

[0130] In this embodiment, the real-time behavior data extraction identifier is a unique data identifier corresponding to the user's real-time behavior data. A matching query is performed in the database through the real-time behavior data extraction identifier, thereby effectively extracting the user's real-time behavior data.

[0131] Step S602, performing behavior analysis on the user's real-time behavior data to obtain user behavior information;

[0132] In this embodiment, the steps of behavior analysis include behavior classification: classifying user behaviors by type, such as browsing, clicking, adding to shopping cart, purchasing, etc.; behavior sequence analysis: analyzing the sequence and frequency of user behaviors to identify user behavior patterns and points of interest; preference extraction: extracting key information such as user product preferences, price sensitivity, and brand preferences based on user behavior data. User behavior information includes basic behavior information: browsing behavior: the behavior of users browsing different pages, products, content, etc. on the website, including browsing time, page dwell time, click-through rate, etc.; search behavior: the behavior of users searching for specific keywords or content in search engines or platforms, reflecting the interests and needs of users; click behavior: the behavior of users clicking on links, buttons, advertisements and other elements, which is an important way for users to interact with page content. Transaction behavior information: purchase behavior: the behavior of users purchasing goods or services on e-commerce platforms, including the types, quantities, and amounts of purchased goods; payment behavior: the behavior of users selecting payment methods and completing payment processes, reflecting the payment habits and preferences of users; return behavior: the behavior of users returning goods after purchase due to dissatisfaction with the goods, which is an important indicator for enterprises to evaluate product quality and user satisfaction. Interactive behavior information: Comment behavior: users' evaluation behavior on products, services or content, including text comments, ratings, etc., reflects users' satisfaction and feedback; Sharing behavior: users' sharing of content, goods or information with friends or on social media helps expand brand influence; Like and collection behavior: users' like and collection behavior on content or goods indicate users' recognition of the content or potential purchase intention.

[0133] Step S603, performing product matching, priority sorting, and group optimization processing on the recommended product group according to the user behavior information to obtain the effective recommended product group.

[0134] In this embodiment, product matching refers to selecting the products that best match the user from the recommended product group based on the user's behavior preferences and interests; priority sorting: prioritizing the products in the recommended product group based on the weight of the user's behavior (such as the weight of purchase behavior is higher than that of browsing behavior) and the relevance of the products; group optimization: considering the diversity and complementarity of the products in the recommended product group, grouping and optimizing the products to ensure that each group can cover the different needs and interests of the user. By matching, prioritizing, and optimizing the recommended product group based on user behavior information, the effective recommended product group that is finally displayed can be further optimized based on real-time user behavior information to improve user browsing satisfaction.

[0135] This embodiment obtains a real-time behavior data extraction identifier, extracts the user's real-time behavior data from a database according to the real-time behavior data extraction identifier; performs behavior analysis on the user's real-time behavior data to obtain user behavior information; and performs product matching, priority sorting, and group optimization processing on the recommended product group according to the user behavior information, thereby obtaining a diverse and effective recommended product group that meets the user's behavior intentions, thereby improving the user's browsing satisfaction and further improving the user's conversion rate.

[0136] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0137] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0138] Further references Fig. 9 , as a response to the above Figure 1 In order to realize the method shown in the figure, the present application provides an embodiment of a data recommendation device, and the device embodiment is Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0139] like Fig. 9 As shown, the data recommendation device 700 described in this embodiment includes: a data acquisition module 701, a data conversion module 702, a data matching module 703, a data evaluation module 704, a data classification module 705, and a data recommendation module 706. Among them:

[0140] The data acquisition module 701 is used to acquire commodity transaction logs and extract commodity metadata and user historical behavior data from the commodity transaction logs;

[0141] A data conversion module 702 is used to vectorize the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector;

[0142] A data matching module 703 is used to perform similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set;

[0143] The data evaluation module 704 is used to extract features from the recommended product set to obtain recommended product attribute features, and input the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score;

[0144] A data classification module 705 is used to generate a recommended product list according to the recommended product scores, and cluster the recommended product list to obtain a recommended product group;

[0145] The data recommendation module 706 is used to obtain user real-time behavior data, adjust the recommended product group according to the user real-time behavior data to obtain a valid recommended product group, and perform data recommendation based on the valid recommended product group.

[0146] By adopting the above-mentioned data recommendation device, this embodiment can obtain commodity transaction logs, extract commodity metadata and user historical behavior data from the commodity transaction logs; vectorize the commodity metadata and the user historical behavior data to obtain commodity feature vectors and user feature vectors; perform similarity matching based on the commodity feature vectors and the user feature vectors to obtain a recommended commodity set; extract features from the recommended commodity set to obtain recommended commodity attribute features, and input the recommended commodity attribute features into a pre-trained commodity evaluation model to obtain recommended commodity scores; generate a recommended commodity list based on the recommended commodity scores, and cluster and group the recommended commodity list to obtain a recommended commodity group; obtain user real-time behavior data, adjust the recommended commodity group based on the user real-time behavior data to obtain an effective recommended commodity group, and perform data recommendation based on the effective recommended commodity group. In this way, accurate and personalized data recommendations can be effectively made to users to improve the user conversion rate.

[0147] To solve the above technical problems, the present application also provides a computer device. Fig.10 , Fig.10 This is a basic structural block diagram of the computer device in this embodiment.

[0148] The computer device 8 includes a memory 81, a processor 82, and a network interface 83 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 8 with components 81-83, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0149] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0150] The memory 81 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 81 can be an internal storage unit of the computer device 8, such as a hard disk or memory of the computer device 8. In other embodiments, the memory 81 can also be an external storage device of the computer device 8, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 8. Of course, the memory 81 can also include both the internal storage unit of the computer device 8 and its external storage device. In this embodiment, the memory 81 is generally used to store the operating system and various application software installed on the computer device 8, such as computer-readable instructions of the data recommendation method, etc. In addition, the memory 81 can also be used to temporarily store various types of data that have been output or are to be output.

[0151] The processor 82 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 82 is generally used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to run computer-readable instructions stored in the memory 81 or process data, such as computer-readable instructions for running the data recommendation method.

[0152] The network interface 83 may include a wireless network interface or a wired network interface. The network interface 83 is generally used to establish a communication connection between the computer device 8 and other electronic devices.

[0153] By adopting the above-mentioned computer device, this embodiment can obtain commodity transaction logs, extract commodity metadata and user historical behavior data from the commodity transaction logs; vectorize the commodity metadata and the user historical behavior data to obtain commodity feature vectors and user feature vectors; perform similarity matching based on the commodity feature vectors and the user feature vectors to obtain a recommended commodity set; extract features from the recommended commodity set to obtain recommended commodity attribute features, and input the recommended commodity attribute features into a pre-trained commodity evaluation model to obtain recommended commodity scores; generate a recommended commodity list based on the recommended commodity scores, and cluster and group the recommended commodity list to obtain a recommended commodity group; obtain user real-time behavior data, adjust the recommended commodity group based on the user real-time behavior data to obtain an effective recommended commodity group, and perform data recommendation based on the effective recommended commodity group. In this way, accurate and personalized data recommendations can be effectively made to users to improve the user conversion rate.

[0154] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the data recommendation method as described above.

[0155] By adopting the above-mentioned computer-readable storage medium, this embodiment can obtain commodity transaction logs, extract commodity metadata and user historical behavior data from the commodity transaction logs; vectorize the commodity metadata and the user historical behavior data to obtain commodity feature vectors and user feature vectors; perform similarity matching based on the commodity feature vectors and the user feature vectors to obtain a recommended commodity set; extract features from the recommended commodity set to obtain recommended commodity attribute features, and input the recommended commodity attribute features into a pre-trained commodity evaluation model to obtain recommended commodity scores; generate a recommended commodity list based on the recommended commodity scores, and cluster and group the recommended commodity list to obtain a recommended commodity group; obtain user real-time behavior data, adjust the recommended commodity group based on the user real-time behavior data to obtain an effective recommended commodity group, and perform data recommendation based on the effective recommended commodity group. In this way, accurate and personalized data recommendations can be effectively made to users to improve user conversion rates.

[0156] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an 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, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0157] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

[0158] The non-Company software tools or components appearing in the embodiments of this application are merely examples and do not represent actual use.

Claims

1. A data recommendation method, characterized in that: The steps include: Obtaining commodity transaction logs, and extracting commodity metadata and user historical behavior data from the commodity transaction logs; Vectorizing the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector; Performing similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set; Extracting features from the recommended product set to obtain recommended product attribute features, and inputting the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score; Generating a recommended product list according to the recommended product scores, and clustering and grouping the recommended product list to obtain a recommended product group; Real-time user behavior data is obtained, the recommended product group is adjusted according to the real-time user behavior data to obtain a valid recommended product group, and data recommendation is performed according to the valid recommended product group.

2. The data recommendation method according to claim 1, characterized in that: The step of obtaining the commodity transaction log and extracting commodity metadata and user historical behavior data from the commodity transaction log specifically includes: Get the transaction log extraction identifier; Extracting the commodity transaction log from the database according to the transaction log extraction identifier; The commodity transaction log is parsed to obtain the commodity metadata and the user historical behavior data.

3. The data recommendation method according to claim 1, characterized in that: The step of vectorizing the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector specifically includes: Extracting features from the product metadata and the user historical behavior data to obtain product interaction features and user historical behavior features; Performing a first dimensionality reduction process on the commodity interaction feature to obtain the commodity feature vector; A second dimensionality reduction process is performed on the user historical behavior features to obtain the user feature vector.

4. The data recommendation method according to claim 1, characterized in that: The step of performing similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set specifically includes: Performing cosine similarity calculation on the product feature vector based on a collaborative filtering algorithm to obtain product similarity, and generating a first recommended product based on the product similarity; Performing cosine similarity calculation on the user feature vector based on a collaborative filtering algorithm to obtain user similarity, and generating a second recommended product according to the user similarity; The data of the first recommended product and the second recommended product are integrated to obtain the recommended product set.

5. The data recommendation method according to claim 1, characterized in that: The step of extracting features from the recommended product set to obtain recommended product attribute features, and inputting the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score specifically includes: Preprocessing the recommended product set to obtain a standard recommended product set; Acquire commodity attribute information, and perform feature extraction on the standard recommended commodity set according to the commodity attribute information to obtain attribute features of the recommended commodity; A model extraction identifier is obtained, the product evaluation model is extracted from a database according to the model extraction identifier, and the attribute characteristics of the recommended product are input into the product evaluation model to obtain the recommended product score.

6. The data recommendation method according to claim 1, characterized in that: The step of generating a recommended product list according to the recommended product scores, and clustering and grouping the recommended product list to obtain a recommended product group specifically includes: Sort the products according to the recommended product scores to obtain a recommended product ranking table; Obtaining a recommendation score threshold, and filtering the recommended product ranking table according to the recommendation score threshold to obtain the recommended product list; Clustering and grouping the recommended product list according to a clustering algorithm to obtain a recommended product cluster; Obtaining category information of the recommended product cluster, and obtaining a category sorting strategy according to the category information; The recommended product clusters are sorted and screened according to the category sorting strategy to obtain the recommended product group.

7. The data recommendation method according to claim 1, characterized in that: The step of obtaining the user's real-time behavior data and adjusting the recommended product group according to the user's real-time behavior data to obtain a valid recommended product group specifically includes: Acquire a real-time behavior data extraction identifier, and extract the user's real-time behavior data from a database according to the real-time behavior data extraction identifier; Performing behavior analysis on the real-time behavior data of the user to obtain user behavior information; The recommended product group is matched with products, prioritized, and optimized according to the user behavior information to obtain the effective recommended product group.

8. A data recommendation device, characterized in that: include: A data acquisition module, used to acquire commodity transaction logs, and extract commodity metadata and user historical behavior data from the commodity transaction logs; A data conversion module, used to vectorize the product metadata and the user historical behavior data to obtain a product feature vector and a user feature vector; A data matching module, used to perform similarity matching based on the product feature vector and the user feature vector to obtain a recommended product set; A data evaluation module is used to extract features from the recommended product set to obtain recommended product attribute features, and input the recommended product attribute features into a pre-trained product evaluation model to obtain a recommended product score; A data classification module, used to generate a recommended product list according to the recommended product scores, and cluster the recommended product list to obtain a recommended product group; The data recommendation module is used to obtain real-time user behavior data, adjust the recommended product group according to the real-time user behavior data to obtain a valid recommended product group, and perform data recommendation based on the valid recommended product group.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the data recommendation method according to any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the data recommendation method according to any one of claims 1 to 7 are implemented.

Images