Product data processing method, device, equipment and storage medium

Abstract

The present invention relates to the field of big data technology, and provides a method, device, equipment and storage medium for processing product data, which are used to improve the accuracy of product data processed based on intelligent recommendation. The method for processing product data includes: classifying the product data to be processed according to the data type to obtain initial text data and initial numerical data; semantically translating and semantically classifying the initial text data through a language model to obtain candidate text data, and numerically processing the initial numerical data through an artificial neural network model to obtain candidate numerical data; classifying and labeling the candidate text data and the candidate numerical data, wherein the labeled and summarized data include product attributes and customer positioning information; according to preset attribute categories, the labeled and summarized data are stored in a preset product database. In addition, the present invention also relates to blockchain technology, and the product data to be processed can be stored in the blockchain.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a method, device, equipment and storage medium for processing product data. Background Art

[0002] With the development of Internet technology and business, intelligent recommendation of business products has attracted much attention, and the application and management of product data is an important part of intelligent recommendation technology for business products. At present, the application and management of product data generally involves classifying product data and replacing attributes through user tags.

[0003] However, in the above method, the classification method of product data is single, and there is no systematic analysis and planning of product data, which leads to low accuracy of product data after processing based on intelligent recommendation. Summary of the invention

[0004] The present invention provides a product data processing method, device, equipment and storage medium, which are used to improve the accuracy of product data after intelligent recommendation processing.

[0005] A first aspect of the present invention provides a method for processing product data, comprising:

[0006] Acquire product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data;

[0007] By using a preset language model, the initial text data is semantically escaped and semantically classified to obtain candidate text data, and by using a preset artificial neural network model, the initial numerical data is numerically processed to obtain candidate numerical data;

[0008] Performing classification type labeling and aggregation processing on the candidate text data and the candidate numerical data, wherein the labeled and aggregated data includes product attributes and customer group positioning information;

[0009] According to the preset attribute categories, the marked and aggregated data are stored in a preset product database to obtain target product data.

[0010] Optionally, in a first implementation of the first aspect of the present invention, the step of performing classification type marking and aggregation processing on the candidate text data and the candidate numerical data includes:

[0011] By using a preset annotation algorithm, the candidate text data is annotated with product types, and the candidate numerical data is annotated with numerical values ​​and annotated with numerical values ​​to obtain data to be summarized;

[0012] Acquire customer information of the product data to be processed, and classify the data to be aggregated according to the customer information to obtain aggregated data;

[0013] The summarized data is converted into a high-density subgraph.

[0014] Optionally, in a second implementation of the first aspect of the present invention, storing the marked and aggregated data into a preset product database according to preset attribute categories to obtain target product data includes:

[0015] Classifying the labeled and summarized data according to preset attribute categories to obtain classified data;

[0016] Writing the classified data into a preset hash table to obtain data to be stored, and performing slicing processing on the data to be stored to obtain slicing data;

[0017] An index of the sliced ​​data is created to obtain index data, and the index data is written into a preset storage file of a preset product database to obtain target product data.

[0018] Optionally, in a third implementation of the first aspect of the present invention, the obtaining of product data to be processed, classifying the product data to be processed according to data type, and obtaining initial text data and initial numerical data, includes:

[0019] Acquire product data to be processed, and perform semantic recognition on the product data to be processed using a preset semantic recognition model to obtain recognition data;

[0020] The preset character-type decision tree and the digital-type decision tree are called according to the data type, and the recognition data are matched and extracted respectively to obtain initial text-type data and initial numerical-type data.

[0021] Optionally, in a fourth implementation of the first aspect of the present invention, the initial numerical data is numerically processed by a preset artificial neural network model to obtain candidate numerical data, including:

[0022] The data type of the initial numerical data is converted into a preset data type through a preset artificial neural network model to obtain converted numerical data;

[0023] Acquire the target field for converting the numerical data, and identify and acquire the field name corresponding to the target field in the preset field table;

[0024] The converted numerical data and the field name are determined as candidate numerical data.

[0025] Optionally, in a fifth implementation of the first aspect of the present invention, after storing the marked and aggregated data into a preset product database according to preset attribute categories and obtaining the target product data, the method further includes:

[0026] Acquire user search information, and match corresponding customer group data from the target product data according to the user search information, wherein the user search information includes user information and / or product scenario information;

[0027] Perform product attribute clustering analysis on the customer group data by using a preset clustering algorithm to obtain target product attributes;

[0028] According to the target product attributes, the preset product data in the preset product database is fully searched and screened to obtain recommended product data.

[0029] Optionally, in a sixth implementation of the first aspect of the present invention, after performing a full search and screening of preset product data in the preset product database according to the target product attribute to obtain recommended product data, the method further includes:

[0030] Obtaining product data of interest based on the recommended product data, updating the preset product database according to the product data of interest, and optimizing the execution process of the target product data.

[0031] A second aspect of the present invention provides a device for processing product data, comprising:

[0032] A classification module is used to obtain product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data;

[0033] A first processing module is used to perform semantic escape and semantic classification on the initial text data through a preset language model to obtain candidate text data, and to perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data;

[0034] A second processing module is used to classify and label the candidate text data and the candidate numerical data, and to summarize the data, wherein the labeled and summarized data includes product attributes and customer group positioning information;

[0035] The storage module is used to store the marked and summarized data into a preset product database according to preset attribute categories to obtain target product data.

[0036] Optionally, in a first implementation manner of the second aspect of the present invention, the second processing module is specifically configured to:

[0037] By using a preset annotation algorithm, the candidate text data is annotated with product types, and the candidate numerical data is annotated with numerical values ​​and annotated with numerical values ​​to obtain data to be summarized;

[0038] Acquire customer information of the product data to be processed, and classify the data to be aggregated according to the customer information to obtain aggregated data;

[0039] The summarized data is converted into a high-density subgraph.

[0040] Optionally, in a second implementation of the second aspect of the present invention, the storage module is specifically used for:

[0041] Classifying the labeled and summarized data according to preset attribute categories to obtain classified data;

[0042] Writing the classified data into a preset hash table to obtain data to be stored, and performing slicing processing on the data to be stored to obtain slicing data;

[0043] An index of the sliced ​​data is created to obtain index data, and the index data is written into a preset storage file of a preset product database to obtain target product data.

[0044] Optionally, in a third implementation of the second aspect of the present invention, the classification module is specifically used to:

[0045] Acquire product data to be processed, and perform semantic recognition on the product data to be processed using a preset semantic recognition model to obtain recognition data;

[0046] The preset character-type decision tree and the digital-type decision tree are called according to the data type, and the recognition data are matched and extracted respectively to obtain initial text-type data and initial numerical-type data.

[0047] Optionally, in a fourth implementation manner of the second aspect of the present invention, the first processing module is specifically configured to:

[0048] The data type of the initial numerical data is converted into a preset data type through a preset artificial neural network model to obtain converted numerical data;

[0049] Acquire the target field for converting the numerical data, and identify and acquire the field name corresponding to the target field in the preset field table;

[0050] The converted numerical data and the field name are determined as candidate numerical data.

[0051] Optionally, in a fifth implementation of the second aspect of the present invention, the device for processing product data further includes:

[0052] A matching module, used to obtain user search information, and match corresponding customer group data from the target product data according to the user search information, wherein the user search information includes user information and / or product scenario information;

[0053] A clustering analysis module, used to perform product attribute clustering analysis on the customer group data by using a preset clustering algorithm to obtain target product attributes;

[0054] The search and screening module is used to search and screen all the preset product data in the preset product database according to the target product attributes to obtain recommended product data.

[0055] Optionally, in a sixth implementation of the second aspect of the present invention, the device for processing product data further includes:

[0056] The update optimization module is used to obtain the product data of interest based on the recommended product data, update the preset product database according to the product data of interest, and optimize the execution process of the target product data.

[0057] The third aspect of the present invention provides a product data processing device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the product data processing device executes the above-mentioned product data processing method.

[0058] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for processing product data.

[0059] In the technical solution provided by the present invention, product data to be processed is obtained, and data type classification is performed on the product data to be processed to obtain initial text data and initial numerical data; semantic translation and semantic classification are performed on the initial text data through a preset language model to obtain candidate text data, and the initial numerical data is numerically processed through a preset artificial neural network model to obtain candidate numerical data; classification type marking and summary processing are performed on the candidate text data and the candidate numerical data to obtain data to be processed, and the data to be processed includes product attributes and customer positioning information; according to preset attribute categories, the data to be processed is stored in a preset product database to obtain target product data. In an embodiment of the present invention, semantic recognition and numerical recognition are performed on the product data to be processed, semantic escape and semantic classification are performed on the initial text data, numerical processing is performed on the initial numerical data, classification type marking and summary processing are performed on the candidate text data and the candidate numerical data, and the data to be processed is stored in a preset product database according to preset attribute categories. It is achieved that the product data to be processed is classified into product types based on text data and numerical data using multiple classification methods, and the product data to be processed is systematically analyzed and planned, thereby improving the accuracy of classification and labeling of the product data to be processed, thereby improving the accuracy of product data processed based on intelligent recommendation. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A schematic diagram of an embodiment of a method for processing product data in an embodiment of the present invention;

[0061] Figure 2 A schematic diagram of another embodiment of a method for processing product data in an embodiment of the present invention;

[0062] Figure 3 A schematic diagram of an embodiment of a device for processing product data in an embodiment of the present invention;

[0063] Figure 4 A schematic diagram of another embodiment of a device for processing product data in an embodiment of the present invention;

[0064] Figure 5 The figure is a schematic diagram of an embodiment of a device for processing product data in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] Embodiments of the present invention provide a method, device, equipment and storage medium for processing product data, which improve the accuracy of product data processed based on intelligent recommendation.

[0066] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0067] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of a method for processing product data in an embodiment of the present invention includes:

[0068] 101. Obtain product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data.

[0069] It is understandable that the execution subject of the present invention may be a product data processing device, or a terminal or a server, which is not specifically limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example.

[0070] The product data to be processed include the income data, family information and personal information of customers who have been interested in a certain type of product in the past, the current sales data of the product, and the information on activities and customer conversion rate of the promotion products, etc. Sales data such as repurchase status data, customer conversion data, profit share data and combined sales data, etc. Customer personal information such as zodiac sign, constellation, gender, credit level and customer registration, etc. The product data to be processed may correspond to one type of product or multiple types of products. For example, the product data to be processed may be the customer income data, family information and personal information of category A products, the current sales data of the product, and the information on activities and customer conversion rate of the promotion products, etc. The product data to be processed may also be the customer income data, family information and personal information of category A products and category B products respectively, the current sales data of the product, and the information on activities and customer conversion rate of the promotion products, etc.

[0071] The server calls a preset web crawler or crawler to crawl the initial product data on the platform web page, and extracts the stored historical product data from the preset database, performs data cleaning, data reduction and data conversion on the initial product data and historical product data to obtain the product data to be processed. The server performs semantic recognition and numerical recognition on the product data to be processed according to the data types of strings and numbers through a preset recognition model to obtain a recognition result, and classifies the product data to be processed into initial text data (corresponding to the data type of the string) and initial numerical data (corresponding to the data type of the number) according to the recognition result. The recognition model includes algorithms and network structures for string recognition and numerical recognition, wherein the server can also perform string fuzzy matching on the product data to be processed through a preset regular expression to obtain the corresponding initial text data.

[0072] 102. Perform semantic translation and semantic classification on the initial text data through a preset language model to obtain candidate text data, and perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data.

[0073] The server uses the preset language model BERT to perform feature extraction, word vector conversion, statistical analysis and classification on the initial text data in turn to obtain candidate text data. For example, the extracted features include the product's historical sales characteristics, the customer situation characteristics of the product, the product's attribute characteristics, the company's personnel length characteristics, sales method characteristics and purchase method characteristics. The customer situation characteristics include the customer's zodiac characteristics, constellation characteristics, age characteristics and gender characteristics. The statistical analysis includes the proportion of Leos and Goats who purchased the product, the proportion of sales personnel with a length of service of 0-6 months and the proportion of sales by senior personnel, etc. The classification can be classified according to the type of statistical analysis results.

[0074] The server uses a feedback neural network model in a preset artificial neural network (ANN) model to perform feature extraction, feature scaling, feature characterization, outlier identification and feature discretization on the initial numerical data to obtain numerical features, and performs prediction, statistical analysis and numerical classification on the numerical features to obtain candidate numerical data. For example, the prediction and statistical analysis are of activity conversion rate data, product subsidy rate data, commission rate data and management income ratio data; the numerical classification can be classified according to the result type of prediction and statistical analysis.

[0075] 103. Classify and label the candidate text data and the candidate numerical data, and perform aggregation processing, wherein the labeled and aggregated data includes product attributes and customer group positioning information.

[0076] The server annotates the candidate text data according to the type of statistical analysis through the preset annotation algorithm and annotation tool to obtain the first data to be aggregated and processed, for example: the candidate text data is annotated with the customer's zodiac sign, the proportion of the purchased zodiac sign, and the proportion of the salesperson. The server annotates the candidate numerical data according to the type of numerical classification through the preset annotation algorithm and annotation tool to obtain the second data to be aggregated and processed, for example: the candidate numerical data is annotated with the activity conversion rate data, the product subsidy rate data, the commission rate data, and the management income ratio data. A user portrait is created based on the first data to be aggregated and processed, the second data to be aggregated and processed, and the corresponding customer information (to achieve aggregate processing), and the labeled aggregated data including product attributes and customer group positioning information is obtained, wherein the customer group positioning information refers to the information on the principles of treating customers, and the customer group positioning information includes but is not limited to the analysis information of the consumption capacity of the customer group, the trend information of the customer group's consumption products, and the consumption demand information of the customer group.

[0077] 104. According to the preset attribute categories, the marked and aggregated data is stored in a preset product database to obtain target product data.

[0078] The server classifies the marked and summarized data into multiple categories according to the preset attribute categories, writes the multiple categories of data into the preset Excel table respectively, obtains multiple tables, and stores the multiple tables into the preset product database according to the preset storage ratio and storage space, thereby obtaining the target product data. Among them, the preset attribute categories include basic attributes and non-basic attributes, such as: basic attributes such as insurance, loan financing, investment and financial management, credit cards and life, and non-basic attributes such as customer management, company management, channel characteristics, channel area, product management and product association.

[0079] In an embodiment of the present invention, semantic recognition and numerical recognition are performed on the product data to be processed, semantic escape and semantic classification are performed on the initial text data, numerical processing is performed on the initial numerical data, classification type marking and summary processing are performed on the candidate text data and the candidate numerical data, and the data to be processed is stored in a preset product database according to preset attribute categories. It is achieved that the product data to be processed is classified into product types based on text data and numerical data using multiple classification methods, and the product data to be processed is systematically analyzed and planned, thereby improving the accuracy of classification and labeling of the product data to be processed, thereby improving the accuracy of product data processed based on intelligent recommendation.

[0080] See also Figure 2 Another embodiment of the method for processing product data in the embodiment of the present invention includes:

[0081] 201. Obtain product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data.

[0082] Specifically, the server obtains product data to be processed, performs semantic recognition on the product data to be processed through a preset semantic recognition model, and obtains recognition data; calls the preset character decision tree and numeric decision tree according to the data type, matches and extracts the recognition data respectively, and obtains initial text data and initial numerical data.

[0083] The product data to be processed may include but are not limited to image data, voice data and text data. The server extracts the product data to be processed that has undergone data preprocessing from a preset database, and sequentially performs target frame detection, target frame segmentation, target frame image recognition and semantic recognition of the target frame image on the image data in the product data to be processed through the target detection algorithm, image segmentation algorithm, image recognition algorithm and semantic recognition algorithm in the preset semantic recognition model to obtain first data, performs voice recognition and semantic recognition on the voice data in the product data to be processed through the voice recognition algorithm in the semantic recognition model to obtain second data, performs semantic recognition on the text data in the product data to be processed through the hidden Markov model in the semantic recognition model to obtain third data, and determines the first data, the second data and the third data as recognition data.

[0084] The server calls the preset character-type decision tree according to the data type, and performs similarity calculation, threshold judgment and data extraction on the recognition data in turn to obtain the initial text-type data corresponding to the product data to be processed. The server calls the preset digital decision tree according to the data type, and performs similarity calculation, threshold judgment and data extraction on the recognition data in turn to obtain the initial numerical data corresponding to the product data to be processed.

[0085] 202. Perform semantic translation and semantic classification on the initial text data through a preset language model to obtain candidate text data, and perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data.

[0086] The server uses the preset language model BERT to perform feature extraction, word vector conversion, statistical analysis and classification on the initial text data in turn to obtain candidate text data. For example, the extracted features include the product's historical sales characteristics, the customer situation characteristics of the product, the product's attribute characteristics, the company's personnel age characteristics, sales method characteristics and purchase method characteristics. The customer situation characteristics include the customer's zodiac characteristics, constellation characteristics, age characteristics and gender characteristics. The statistical analysis includes the proportion of young people and the proportion of elderly people who purchased the product.

[0087] Specifically, the server converts the data type of the initial numerical data into the preset data type through a preset artificial neural network model to obtain converted numerical data; obtains the target field of the converted numerical data, identifies and obtains the field name corresponding to the target field in the preset field table; and determines the converted numerical data and the field name as candidate numerical data.

[0088] The server uses the self-organizing neural network algorithm in the preset artificial neural network model to extract and classify the features of the initial numerical data to obtain the numerical data to be processed, convert the data type of the numerical data to be processed into the preset data type to achieve the uniform conversion of the data type of the initial numerical data into the same type to obtain the converted numerical data, extract the target field of the converted numerical data, generate the target key of the target field, and match the field name in the preset field table with a key-value pair according to the target key to obtain candidate numerical data, which may include the converted numerical data and the field name corresponding to the converted numerical data. This can achieve systematic analysis and planning of the product data to be processed based on the numerical data, thereby improving the accuracy of the product data processed based on intelligent recommendation.

[0089] 203. Classify and label the candidate text data and the candidate numerical data, and perform aggregation processing, wherein the labeled and aggregated data includes product attributes and customer group positioning information.

[0090] Specifically, the server uses a preset annotation algorithm to annotate the candidate text data with product types, and performs numerical calculation and numerical annotation on the candidate numerical data to obtain the data to be summarized; obtains the customer information of the product data to be processed, and classifies the data to be summarized according to the customer information to obtain the summarized data; and converts the summarized data into a high-density subgraph.

[0091] The server uses preset annotation algorithms and annotation tools (such as BasicFinder) to annotate the candidate text data according to product type. For example, the server annotates the candidate text data with product attributes, product sales type, and product sales customer group type. The server calculates and predicts the candidate numerical data according to the preset annotation numerical rules to obtain the numerical values ​​to be annotated. The server uses the preset annotation algorithms and annotation tools to annotate the numerical values ​​to the candidate numerical data, thereby obtaining the data to be summarized.

[0092] The server obtains the customer information of the product data to be processed and the hash table corresponding to the customer information, classifies the data to be summarized and writes it into the corresponding hash tables to obtain the summarized data, generates a sub-high-density subgraph according to the data in each hash table (i.e., the summarized data), obtains multiple sub-high-density subgraphs, obtains the association relationship between the data in each hash table, connects the multiple sub-high-density subgraphs according to the association relationship, and obtains the main high-density subgraph, i.e., the marked summarized data, so as to facilitate the subsequent rapid retrieval and organization of the product data to be processed.

[0093] 204. According to the preset attribute categories, the marked and aggregated data is stored in a preset product database to obtain target product data.

[0094] Specifically, the server classifies the marked and summarized data according to preset attribute categories to obtain classified data; writes the classified data into a preset hash table to obtain data to be stored, and slices the data to be stored to obtain sliced ​​data; creates an index for the sliced ​​data to obtain index data, and writes the index data to a preset storage file of a preset product database to obtain target product data.

[0095] After the server obtains the classified data, it writes the classified data into the preset hash tables corresponding to the attribute categories to obtain the data to be stored, divides each hash table into multiple sub-tables, that is, slices the data to be stored to obtain sliced ​​data, creates indexes between the multiple sub-tables, obtains the indexes corresponding to the multiple hash tables, the indexes corresponding to the multiple hash tables are related, obtains the index data, and stores the index data in the preset storage file of the preset product database, thereby obtaining the target product data, which improves the retrieval efficiency and accuracy of the target product data.

[0096] 205. Obtain user search information, and match corresponding customer group data from target product data according to the user search information. The user search information includes user information and / or product scenario information.

[0097] The user enters user search information in the preset interface, and the preset interface sends the user search information to the server. The server receives the user search information, extracts key information of the user search information, and searches the target product data in the preset database according to the key information to obtain the corresponding customer group data, which includes the customer group to which the user belongs and the customer product data of the customer group, wherein the user information includes at least one of the user's income information, the province and city where the user lives, the family income information, the family expenditure information, the family member information, the purchased financial product information, and the debt information. Product scenario information, for example: education and parenting, health care, cars and travel, small and micro business and wealth management, etc.

[0098] 206. Perform product attribute clustering analysis on customer group data using a preset clustering algorithm to obtain target product attributes.

[0099] The server performs product attribute clustering analysis on customer group data through a preset clustering algorithm - the expectation maximization algorithm (EM) clustering algorithm based on Gaussian mixture model (adaptive background mixture models for real-time tracking, GMM) to infer product attributes that meet user retrieval information (current status), obtain multiple candidate product attributes, obtain the numerical value marked in the target product data corresponding to each candidate product attribute, perform weighted summation on the numerical value to obtain the target value, and determine the candidate product attribute corresponding to the target value greater than the preset threshold as the target product attribute to improve the accuracy of the target product attribute.

[0100] 207. According to the target product attributes, the preset product data in the preset product database is fully searched and filtered to obtain recommended product data.

[0101] The server calls a preset search engine, and based on the target product attributes, performs a full search of the preset product data in the preset product database to obtain candidate product data, wherein the candidate product data includes product data corresponding to multiple candidate products, and the preset product data includes product attributes and attribute values. The attribute values ​​are normalized to obtain normalized values, and the candidate product data are sorted in descending order of the normalized values ​​to obtain sequence data. The corresponding candidate product data in the sequence data is extracted according to the preset ranking range, thereby obtaining recommended product data.

[0102] By matching the corresponding customer group data from the target product data based on user retrieval information, performing product attribute clustering analysis on the customer group data through a preset clustering algorithm, and conducting a full search and screening of the preset product data in the preset product database based on the target product attributes, the precise matching of the recommended product data that meets user needs is achieved, labor costs are reduced, and the matching efficiency of product data is improved. The recommended product data that meets user needs can be analyzed from a multi-dimensional perspective, and the recommended product data that meets user needs can be deeply mined. On the basis of improving the accuracy of product data processed based on intelligent recommendation, the accuracy of intelligent recommendation of the processed product data is improved.

[0103] Specifically, the server searches and filters all preset product data in the preset product database according to the target product attributes, obtains recommended product data, and then obtains interest product data based on the recommended product data, updates the preset product database according to the interest product data, and optimizes the execution process of the target product data.

[0104] After the server obtains the recommended product data, it renders the recommended product data to the preset display page. The user clicks on the recommended product data of interest on the preset display page, i.e., the product data of interest. The preset display page records and sends the product data of interest to the server. The server receives the product data of interest, writes the relevant fields or field values ​​of the product of interest in the tag content of the product of interest, and replaces the original product data of interest in the preset product database with the product data of interest after writing the tag content, so as to update the preset product database, and optimize and adjust the execution process of the target product data, the optimization data labeling algorithm rules used, and the network structure and algorithm of the model used according to the product of interest data. By updating the preset product database according to the product of interest data and optimizing the execution process of the target product data, the accuracy of the product data processed based on the intelligent recommendation, as well as the retrieval accuracy and retrieval efficiency of the target product data are improved.

[0105] In the embodiments of the present invention, not only is it possible to implement the product type classification based on text data and the numerical data classification for the product data to be processed by using multiple classification methods, but also to perform systematic analysis and planning on the product data to be processed, thereby improving the accuracy of classification and labeling of the product data to be processed, and thus improving the accuracy of the product data processed based on intelligent recommendation, but also by matching the corresponding customer group data from the target product data according to the user retrieval information, performing product attribute clustering analysis on the customer group data through a preset clustering algorithm, and performing a full search and screening of the preset product data in the preset product database according to the target product attributes, thereby achieving accurate matching of the recommended product data required by the user, improving the matching efficiency of the product data, being able to analyze the recommended product data required by the user from multiple dimensions, and deeply mining the recommended product data required by the user, and improving the accuracy of the intelligent recommendation of the processed product data on the basis of improving the accuracy of the product data processed based on the intelligent recommendation.

[0106] The above describes the method for processing product data in the embodiment of the present invention. The following describes the device for processing product data in the embodiment of the present invention. Figure 3 , an embodiment of a device for processing product data in an embodiment of the present invention includes:

[0107] The classification module 301 is used to obtain product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data;

[0108] The first processing module 302 is used to perform semantic translation and semantic classification on the initial text data through a preset language model to obtain candidate text data, and to perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data;

[0109] The second processing module 303 is used to classify and label the candidate text data and the candidate numerical data, and to summarize the labeled and summarized data, wherein the labeled and summarized data includes product attributes and customer group positioning information;

[0110] The storage module 304 is used to store the marked and summarized data into a preset product database according to preset attribute categories to obtain target product data.

[0111] The functional implementation of each module in the above-mentioned product data processing device corresponds to each step in the above-mentioned product data processing method embodiment, and its functions and implementation processes are no longer described one by one here.

[0112] In an embodiment of the present invention, semantic recognition and numerical recognition are performed on the product data to be processed, semantic escape and semantic classification are performed on the initial text data, numerical processing is performed on the initial numerical data, classification type marking and summary processing are performed on the candidate text data and the candidate numerical data, and the data to be processed is stored in a preset product database according to preset attribute categories. It is achieved that the product data to be processed is classified into product types based on text data and numerical data using multiple classification methods, and the product data to be processed is systematically analyzed and planned, thereby improving the accuracy of classification and labeling of the product data to be processed, thereby improving the accuracy of product data processed based on intelligent recommendation.

[0113] See also Figure 4 Another embodiment of the device for processing product data in the embodiment of the present invention includes:

[0114] The classification module 301 is used to obtain product data to be processed, classify the product data to be processed according to data type, and obtain initial text data and initial numerical data;

[0115] The first processing module 302 is used to perform semantic translation and semantic classification on the initial text data through a preset language model to obtain candidate text data, and to perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data;

[0116] The second processing module 303 is used to classify and label the candidate text data and the candidate numerical data, and to summarize the labeled and summarized data, wherein the labeled and summarized data includes product attributes and customer group positioning information;

[0117] The storage module 304 is used to store the marked and summarized data into a preset product database according to the preset attribute categories to obtain the target product data;

[0118] A matching module 305 is used to obtain user search information and match corresponding customer group data from target product data according to the user search information, wherein the user search information includes user information and / or product scenario information;

[0119] The cluster analysis module 306 is used to perform product attribute cluster analysis on customer group data by using a preset clustering algorithm to obtain target product attributes;

[0120] The search and screening module 307 is used to search and screen all the preset product data in the preset product database according to the target product attributes to obtain recommended product data.

[0121] Optionally, the second processing module 303 may also be specifically configured to:

[0122] Through the preset annotation algorithm, the candidate text data is annotated with product types, and the candidate numerical data is annotated with numerical calculation and numerical annotation to obtain the data to be summarized;

[0123] Obtain customer information of product data to be processed, classify the data to be aggregated according to the customer information, and obtain aggregated data;

[0124] Convert summary data into dense subgraphs.

[0125] Optionally, the storage module 304 may also be specifically used for:

[0126] According to the preset attribute categories, the marked and summarized data are classified to obtain classified data;

[0127] Writing the classified data into a preset hash table to obtain data to be stored, and sharding the data to be stored to obtain sharded data;

[0128] An index of the shard data is created to obtain index data, and the index data is written to a preset storage file of a preset product database to obtain target product data.

[0129] Optionally, the classification module 301 may also be specifically used for:

[0130] Obtain product data to be processed, and perform semantic recognition on the product data to be processed through a preset semantic recognition model to obtain recognition data;

[0131] The preset character decision tree and digital decision tree are called according to the data type to match and extract the recognition data respectively to obtain the initial text data and the initial numerical data.

[0132] Optionally, the first processing module 302 may also be specifically configured to:

[0133] The data type of the initial numerical data is converted into a preset data type through a preset artificial neural network model to obtain converted numerical data;

[0134] Get the target field for converting numeric data, identify and get the field name corresponding to the target field in the preset field table;

[0135] The converted numeric data and field names are determined as candidate numeric data.

[0136] Optionally, the product data processing device further includes:

[0137] The update optimization module 308 is used to obtain the product data of interest based on the recommended product data, update the preset product database according to the product data of interest, and optimize the execution process of the target product data.

[0138] The functional implementation of each module and each unit in the above-mentioned product data processing device corresponds to each step in the above-mentioned product data processing method embodiment, and their functions and implementation processes are no longer described one by one here.

[0139] In the embodiments of the present invention, not only is it possible to implement the product type classification based on text data and the numerical data classification for the product data to be processed by using multiple classification methods, but also to perform systematic analysis and planning on the product data to be processed, thereby improving the accuracy of classification and labeling of the product data to be processed, and thus improving the accuracy of the product data processed based on intelligent recommendation, but also by matching the corresponding customer group data from the target product data according to the user retrieval information, performing product attribute clustering analysis on the customer group data through a preset clustering algorithm, and performing a full search and screening of the preset product data in the preset product database according to the target product attributes, thereby achieving accurate matching of the recommended product data required by the user, improving the matching efficiency of the product data, being able to analyze the recommended product data required by the user from multiple dimensions, and deeply mining the recommended product data required by the user, and improving the accuracy of the intelligent recommendation of the processed product data on the basis of improving the accuracy of the product data processed based on the intelligent recommendation.

[0140] above Figure 3 and Figure 4The device for processing product data in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The device for processing product data in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0141] Figure 5 1 is a schematic diagram of the structure of a product data processing device provided by an embodiment of the present invention. The product data processing device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 can be short-term storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the product data processing device 500. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the product data processing device 500.

[0142] The product data processing device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 5 The structure of the product data processing device shown does not constitute a limitation on the product data processing device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0143] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the product data processing method.

[0144] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.

[0145] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0148] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for processing product data, characterized in that: The method for processing product data includes: Acquire product data to be processed, classify the product data to be processed according to data types, and obtain initial text data and initial numerical data; wherein, classifying the product data to be processed according to data types to obtain initial text data and initial numerical data includes: using a preset recognition model, according to the data types of character strings and numerical values, performing semantic recognition and numerical recognition on the product data to be processed to obtain recognition results, and classifying the product data to be processed into initial text data and initial numerical data according to the recognition results, wherein the recognition model includes algorithms and network structures for character string recognition and numerical recognition; By using a preset language model, the initial text data is semantically escaped and semantically classified to obtain candidate text data, and by using a preset artificial neural network model, the initial numerical data is numerically processed to obtain candidate numerical data; Performing classification type labeling and aggregation processing on the candidate text data and the candidate numerical data, wherein the labeled and aggregated data includes product attributes and customer group positioning information; The step of classifying and summarizing the candidate text data and the candidate numerical data includes: By using a preset annotation algorithm, the candidate text data is annotated with product types, and the candidate numerical data is annotated with numerical values ​​and annotated with numerical values ​​to obtain data to be summarized; Acquire customer information of the product data to be processed, and classify the data to be aggregated according to the customer information to obtain aggregated data; converting the summary data into a high-density subgraph; According to the preset attribute categories, the marked and aggregated data is stored in a preset product database to obtain target product data; The step of storing the marked and aggregated data into a preset product database according to preset attribute categories to obtain target product data includes: Classifying the labeled and summarized data according to preset attribute categories to obtain classified data; Writing the classified data into a preset hash table to obtain data to be stored, and performing slicing processing on the data to be stored to obtain slicing data; An index of the sliced ​​data is created to obtain index data, and the index data is written into a preset storage file of a preset product database to obtain target product data.

2. The method for processing product data according to claim 1, characterized in that: The step of obtaining the product data to be processed and classifying the product data to be processed according to data types to obtain initial text data and initial numerical data includes: Acquire product data to be processed, and perform semantic recognition on the product data to be processed using a preset semantic recognition model to obtain recognition data; The preset character-type decision tree and the digital-type decision tree are called according to the data type, and the recognition data are matched and extracted respectively to obtain initial text-type data and initial numerical-type data.

3. The method for processing product data according to claim 1, characterized in that: The method of numerically processing the initial numerical data by using a preset artificial neural network model to obtain candidate numerical data includes: The data type of the initial numerical data is converted into a preset data type through a preset artificial neural network model to obtain converted numerical data; Acquire the target field for converting the numerical data, and identify and acquire the field name corresponding to the target field in the preset field table; The converted numerical data and the field name are determined as candidate numerical data.

4. The method for processing product data according to any one of claims 1 to 3, characterized in that: After storing the marked and aggregated data into a preset product database according to the preset attribute categories and obtaining the target product data, the method further includes: Acquire user search information, and match corresponding customer group data from the target product data according to the user search information, wherein the user search information includes user information and / or product scenario information; Perform product attribute clustering analysis on the customer group data by using a preset clustering algorithm to obtain target product attributes; According to the target product attributes, the preset product data in the preset product database is fully searched and screened to obtain recommended product data.

5. The method for processing product data according to claim 4, characterized in that: After the preset product data in the preset product database is fully searched and screened according to the target product attributes to obtain the recommended product data, the method further includes: Obtaining product data of interest based on the recommended product data, updating the preset product database according to the product data of interest, and optimizing the execution process of the target product data.

6. A product data processing device, characterized in that: The product data processing device comprises: A classification module, used to obtain product data to be processed, classify the product data to be processed according to data types, and obtain initial text data and initial numerical data; wherein, the classification of the product data to be processed according to data types to obtain initial text data and initial numerical data includes: using a preset recognition model, according to the data types of character strings and numerical values, performing semantic recognition and numerical recognition on the product data to be processed to obtain recognition results, and classifying the product data to be processed into initial text data and initial numerical data according to the recognition results, wherein the recognition model includes algorithms and network structures for character string recognition and numerical recognition; A first processing module is used to perform semantic escape and semantic classification on the initial text data through a preset language model to obtain candidate text data, and to perform numerical processing on the initial numerical data through a preset artificial neural network model to obtain candidate numerical data; The second processing module is used to classify and label the candidate text data and the candidate numerical data and perform summary processing, wherein the labeled and summarized data includes product attributes and customer positioning information; wherein the classification and labeling and summary processing of the candidate text data and the candidate numerical data includes: labeling the candidate text data by product type through a preset labeling algorithm, and performing labeling value calculation and value labeling on the candidate numerical data to obtain data to be summarized; obtaining customer information of the product data to be processed, and classifying the data to be summarized according to the customer information to obtain summary data; converting the summary data into a high-density subgraph; A storage module is used to store the data after the tags are summarized into a preset product database according to preset attribute categories to obtain target product data; wherein, storing the data after the tags are summarized into a preset product database according to preset attribute categories to obtain target product data includes: classifying the data after the tags are summarized according to preset attribute categories to obtain classified data; writing the classified data into a preset hash table to obtain data to be stored, and slicing the data to be stored to obtain slicing data; creating an index of the slicing data to obtain index data, and writing the index data to a preset storage file of the preset product database to obtain target product data.

7. A product data processing device, characterized in that: The product data processing device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the product data processing device to execute the product data processing method according to any one of claims 1 to 5.

8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the method for processing product data as described in any one of claims 1 to 5 is implemented.

Images