An image-based export sales order generation method, device, equipment and storage medium

Abstract

The present application relates to the field of artificial intelligence, and proposes a kind of export sales order generation method, device, equipment and storage medium based on image, by obtaining the consumption image data of export sales object, based on consumption image data, using evaluation model to evaluate the consumption level of export sales object, determine target consumption level, obtain the product price information of product to be sold, and combine the preset price level division rule, determine the product price level of product to be sold, according to the target consumption level of export sales object and product price level, determine the first offer weight, according to the first offer weight and product price information, generate initial product offer, input initial product offer and trade image data into bill generation model, output export sales order for export sales object, according to the consumption level of export sales object, the price level of product to be sold by user, and corresponding regional factor, the offer information of corresponding product to be sold is obtained, increase the rationality of product offer.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, and in particular to a portrait-based export sales order generation method and device, equipment and storage medium. BACKGROUND

[0002] At present, for product pricing, in national or domestic trade, customers inquire about the price of goods from the sales, and the sales report reasonable prices by considering the cost, profit, market competitiveness and other factors of their own products. Reasonable pricing can quickly attract customer attention and increase the probability of closing the order, and ensure product profits. Overly high pricing may deter customers. Low insurance cannot guarantee profits or make some users question product quality. However, when pricing products for export, factors such as the consumption level of the export target, trade data of the export product, such as the currency and tax rate information of the region where the export target is located and the local product to be sold, and the location information of the region where the export target is located, make it difficult to get a reasonable export product pricing sheet. Therefore, how to get a reasonable pricing sheet for export products has become a problem to be solved. SUMMARY

[0003] Therefore, it is necessary to provide a portrait-based export sales order generation method, device, equipment and storage medium to solve the problem of how to get a reasonable pricing sheet for export products.

[0004] In a first aspect, a portrait-based export sales order generation method is provided, the method comprising:

[0005] Using a deep learning algorithm to perform data relationship mining processing on portrait data of historical export targets, obtaining consumption portrait related factors of the historical export targets, and constructing an evaluation model based on the consumption portrait related factors; the grade evaluation model is used to evaluate the consumption grade of the export target;

[0006] Obtain the consumption portrait data of the export target, use the evaluation model to evaluate the consumption grade of the export target based on the consumption portrait data, and determine the evaluation result as the target consumption grade of the export target;

[0007] Obtain product price information of the product to be sold, and determine the product price grade of the product to be sold based on a preset price grade division rule. According to the target consumption grade of the export target and the product price grade, determine the first pricing weight, and according to the first pricing weight and the product price information, generate an initial product pricing;

[0008] obtain trade portrait data of the export object, the trade portrait data including tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and the product to be sold is originally from, and location information of the region where the export object is located;

[0009] input the initial product offer and the trade portrait data into a bill generation model, and output an export bill for the export object.

[0010] In a second aspect, an export bill generation device based on a portrait is provided, and the device includes:

[0011] a mining module configured to perform data relationship mining processing on portrait data of historical export objects by using a deep learning algorithm, obtain a consumption portrait related factor of the historical export objects, and construct an evaluation model based on the consumption portrait related factor; and the evaluation model is configured to evaluate a consumption level of the export object;

[0012] a consumption level evaluation module configured to obtain consumption portrait data of the export object, evaluate a consumption level of the export object by using the evaluation model based on the consumption portrait data, and determine an evaluation result as a target consumption level of the export object;

[0013] an initial product offer module configured to obtain product price information of a product to be sold, determine a product price level of the product to be sold by combining a preset price level division rule, determine a first offer weight according to the target consumption level of the export object and the product price level, and generate an initial product offer according to the first offer weight and the product price information;

[0014] a trade portrait data obtaining module configured to obtain trade portrait data of the export object, the trade portrait data including tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and the product to be sold is originally from, and location information of the region where the export object is located;

[0015] an export bill generation module configured to input the initial product offer and the trade portrait data into a bill generation model, and output an export bill for the export object.

[0016] In a third aspect, a computer device is provided, and the computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, and the processor implements the export bill generation method based on a portrait when executing the computer program.

[0017] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium storing a computer program, the computer program being executed by a processor to implement the export sales order generation method based on portraits according to the first aspect.

[0018] Compared with the prior art, the present application has the following beneficial effects:

[0019] The historical export object portrait data is processed by using a deep learning algorithm to mine data relationships, to obtain consumption portrait related factors of the historical export object, and an evaluation model is constructed based on the consumption portrait related factors, the evaluation model being used to evaluate the consumption level of the export object, product price information of the product to be sold is obtained, and a preset price level division rule is combined to determine the product price level of the product to be sold, a first offer weight is determined according to the target consumption level of the export object and the product price level, an initial product offer is generated according to the first offer weight and the product price information, trade portrait data of the export object is obtained, the trade portrait data including tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and the product to be sold, and location information of the region where the export object is located, the initial product offer and the trade portrait data are input into a document generation model, and an export sales order for the export object is output, the corresponding product offer information is obtained according to the consumption level of the export object, the price level of the product to be sold, and the corresponding regional factors, and the rationality of the product offer is increased. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the description of the embodiments of the present application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0021] Figure 1 is an application environment schematic diagram of an export sales order generation method based on portraits provided by an embodiment of the present application;

[0022] Figure 2 is a flowchart of an export sales order generation method based on portraits provided by an embodiment of the present application;

[0023] Figure 3 is a flowchart of an export sales order generation method based on portraits provided by an embodiment of the present application;

[0024] Figure 4 is a structural schematic diagram of an export sales order generation device based on portraits provided by an embodiment of the present application;

[0025] Figure 5 is a structural schematic diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort fall within the scope of the present application.

[0027] It should be understood that, when used in the specification and the appended claims of the present application, the term “comprising” indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0028] It should also be understood that, when used in the specification and the appended claims of the present application, the term “and / or” refers to any combination of one or more of the associated listed items and all possible combinations thereof, and includes these combinations.

[0029] As used in the specification and the appended claims of the present application, the term “if” can be interpreted as “when” or “upon” or “in response to a determination” or “in response to detecting” depending on the context. Similarly, the phrase “if it is determined” or “if [a described condition or event] is detected” can be interpreted as meaning “upon determining” or “in response to determining” or “upon detecting [a described condition or event]” or “in response to detecting [a described condition or event]” depending on the context.

[0030] In addition, in the description of the present application and the appended claims, the terms “first”, “second”, “third”, etc. are only used for differentiation in description, and cannot be understood as indicating or implying relative importance.

[0031] In the present specification, the phrase “one embodiment” or “some embodiments” or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Therefore, the occurrences of the phrases “in one embodiment”, “in some embodiments”, “in other some embodiments”, “in yet other embodiments”, and the like in various places in the specification are not necessarily all referring to the same embodiment, unless otherwise specifically noted. The terms “comprise”, “include”, “have”, and their conjugates, mean “including but not limited to”, unless otherwise specifically noted.

[0032] The embodiment of the present application can acquire and process related data based on artificial intelligence technology. The artificial intelligence (AI) is the theory, method, technology and application system for using digital computers or computer-controlled machines to simulate, extend and expand human intelligence, perceive environment, acquire knowledge and use knowledge to obtain optimal results.

[0033] The artificial intelligence basic technology generally includes technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. The artificial intelligence software technology mainly includes computer vision technology, robot technology, biometric identification technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc.

[0034] It should be understood that the size of the serial number of each step in the following embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0035] In order to illustrate the technical solutions of the present application, the following will be illustrated by specific embodiments.

[0036] The portrait-based export sales order generation method provided by the embodiment of the present application can be applied in the application environment such as Figure 1 , wherein the client and the server communicate. The client includes but is not limited to palmtop computer, desktop computer, notebook computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA) and other computer devices. The server can be realized by an independent server or a server cluster composed of multiple servers.

[0037] Referring to Figure 2 , it is a flowchart of a portrait-based export sales order generation method provided by an embodiment of the present application. The portrait-based export sales order generation method can be applied to the server in Figure 1 , and the server is connected to the corresponding client to provide model training service for the client. As shown in Figure 2 , the portrait-based export sales order generation method can include the following steps.

[0038] S201: using a deep learning algorithm to perform data relationship mining processing on portrait data of historical export objects, acquiring consumption portrait related factors of the historical export objects, and constructing an evaluation model based on the consumption portrait related factors.

[0039] In step S201, the introduction of the deep learning algorithm is valuable for big data mining and prediction analysis. Deep learning is a process of combining shallow basic features to mine high-level abstract features in the learning data. Deep learning can effectively improve the experimental results. Data relationship mining processes analysis, calculation, arrangement, and other processing of data to construct an evaluation model

[0040] In this embodiment, the historical export target portrait data is portrait data of different regions, different times, and different scales. The historical export target portrait data is preprocessed and analyzed. Data preprocessing supplements, cleans, and calculates the historical export target portrait data. Data analysis uses a distributed database or a distributed computing cluster to perform general analysis and classification summary on the massive data stored therein to meet most common analysis needs. Data cleaning cleans error data in the database. Because the data in the database is a collection of data related to a certain theme, the data is extracted from multiple business systems and contains historical data. Therefore, some data is error data, and some data conflicts with each other. Therefore, error data and conflict data need to be cleaned. For example, in the sales process, if there is text information in the ID information in the personal basic information, it is considered to be error data, and the ID information data in the personal basic information is cleaned. Data supplementation is to perfect the missing data. For example, in the bank transaction process, the data of each transaction is an important reference for prediction. When the transaction data is missing, it has an impact on future transaction prediction, and complete transaction data needs to be supplemented. Data preprocessing standardizes the data. A large input value may slow down the learning speed, so StandardScaler is used to scale the standardized data. StandardScaler standardizes the data by normalizing the data for each dimension.

[0041] After processing the data, the data features are selected. In this embodiment, the consumption level of the export target is evaluated. Some influencing factors related to the consumption level are selected in combination with the characteristics of the historical export target portrait data. First, the architecture is defined, and the original data is processed using a recurrent neural network. The recurrent neural network model can reduce the training cost of the model, eliminate the redundancy of negative class data samples, automatically learn features at the network bottom layer, and mine more valuable information from the data. According to the learned data features, an evaluation model is constructed. In the recurrent neural network, the Sigmoid function is selected as the activation function. When the number of training increases but the value of the loss function no longer changes with it, the training is stopped, and the corresponding model is output.

[0042] Optionally, the portrait data of the historical export object is processed by using the deep learning algorithm to mine the data relationship, obtain the related consumption portrait data of the historical export object, and construct an evaluation model based on the related consumption portrait data, including:

[0043] The portrait data of the historical export object is processed by using the Apriori algorithm, and the consumption portrait related factors of the evaluation level are obtained according to the correlation between the consumption portrait related factors and the evaluation level in the portrait data of the historical export object.

[0044] The evaluation level consumption portrait related factors are analyzed, the feature engineering of the evaluation level consumption portrait related factors is extracted, and the evaluation model is constructed.

[0045] In this embodiment, the portrait data of the historical export object is processed by using the Apriori algorithm, and the consumption portrait related factors of the historical export object are obtained according to the correlation between the consumption portrait related factors in the portrait data of the historical export object. According to the analysis of the consumption portrait related factors of the portrait data of the historical export object and the extraction of the feature engineering of the consumption portrait related factors of the portrait data of the historical export object, the evaluation model is constructed.

[0046] In this embodiment, in the process of data feature extraction, the portrait data of the historical export object is processed by using the Apriori algorithm according to the consumption portrait related factors of the historical export object as the features of the evaluation model to be constructed, and the consumption portrait related factors of the consumption level are obtained according to the correlation between the consumption portrait related factors in the portrait data of the historical export object. The Apriori algorithm uses the known frequency as the standard to find the frequent item set through association, and the elements in the frequent item set are associated. Therefore, the factors related to the consumption level can be obtained from the frequent item set.

[0047] It should be noted that the evaluation level consumption portrait related factors are selected from different dimensions, and the dimensions and units are not unified, which will affect the weight of the model evaluation features and further affect the prediction effect of the model. Therefore, the feature normalization processing is needed to scale the feature data to a smaller interval range. The commonly used method is Min-Max standardization processing. In the data mining process, the feature extraction is performed to obtain the data features.

[0048] S202: Obtain the consumption portrait data of the export object, use the evaluation model to evaluate the consumption level of the export object based on the consumption portrait data, and determine the evaluation result as the target consumption level of the export object.

[0049] In step S202, the consumption portrait data of the export target includes consumption level information of a region where the export target is located, consumption level information of an industry where the export target is located, and consumption capacity information of the export target, and the evaluation model is a classification model obtained through deep learning.

[0050] In this embodiment, the consumption portrait data of the export target is obtained, including consumption level information of a region where the export target is located, consumption level information of an industry where the export target is located, and consumption capacity information of the export target. When obtaining the consumption portrait data of the export target, the corresponding consumption portrait data of the export target is obtained through a crawler technology. The consumption expenditure data of the export target is searched through a breadth-first search algorithm, which mainly realizes the traversal of a tree graph along the tree width, and the graph algorithm is also relatively traditional. Once the target is found, the algorithm is terminated immediately. The process of designing and implementing the algorithm is relatively simple, and the search is in the category of blind search. In addition, combined with breadth-first search and webpage filtering, webpage crawling is mainly realized by relying on the breadth-first strategy, and then irrelevant webpages are filtered.

[0051] It should be noted that the keywords in the export target can also be used to set the corresponding theme, search the consumption portrait data of the export target, and serve as a theme crawler. The theme crawler starts from a group of seed pages related to the theme, obtains the page information pointed to by the URL, saves the pages related to the set theme, extracts new URL links from these pages, and puts the links into the URL queue after estimating and scoring the value of the links. The scheduling module takes the head URL as the next page to be downloaded. In this embodiment, the crawler theme can be set as the consumption portrait data.

[0052] Based on the consumption portrait data, the evaluation model is used to evaluate the consumption level of the export target, and the evaluation result is the target consumption level of the export target. The evaluation model is a classification model obtained through neural network learning. The essence is to provide the input layer of the neural network with the consumption portrait data information of the export target as input, and after the excitation and transmission between neurons, the weights and thresholds between neurons are continuously adjusted to adapt to the current input vector features, and similar feature inputs in the future produce similar output patterns. In actual consumption level evaluation applications, similar to logistic regression analysis, the neural network will discretize the consumption level of the export target into high-consumption export targets, medium-consumption export targets, and low-consumption export targets.

[0053] It should be noted that the maximum characteristic of the evaluation model based on the neural network is that the self-learning ability of the neural network can be used to fully learn the consumption portrait data of the export target. Finally, a hidden model is obtained and stored in the specific connection structure of the neural network, forming a non-linear mapping function, that is, the consumption level can be mapped according to the characteristics of the consumption portrait data of the export target, and the export target is evaluated.

[0054] Generally, in an artificial neural network, input signals enter an input layer, a hidden layer, and finally an output layer in sequence. According to the transmission direction of the input signals between neurons, it can be divided into forward propagation network and feedback propagation network. In the feedback propagation network, the signals of the output layer continue to propagate to the input layer as the input of the next iteration.

[0055] S203: Obtain product price information of the product to be sold, determine the product price level of the product to be sold in combination with a preset price level division rule, determine the first offer weight according to the target consumption level of the export target and the product price level, and generate an initial product offer according to the first offer weight and the product price information.

[0056] In step S203, the price level of the product to be sold is divided according to the preset price level division rule. When dividing the product price level, different product prices correspond to different price ranges. According to the target consumption level of the export target and the product price level, the first offer weight is determined, and the initial product offer is generated according to the first offer weight and the product price information. The initial product offer reflects the different values of the product to different consumers.

[0057] In this embodiment, the product price is first divided into levels, and the product price is divided into high-price products, medium-price products and low-price products. By clustering the different prices of the product to be sold, the product price range is determined, and the clustering result is divided into different levels.

[0058] The consumption level and the product price level are combined, any consumption level and any price level are combined two by two, different combinations set different values for the corresponding consumption level and price level as the first offer weight, and the first offer weight and the product price information are combined to obtain the initial product price.

[0059] Optionally, obtaining product price information of the product to be sold, and determining the product price level of the product to be sold in combination with a preset price level division rule, comprises:

[0060] Obtaining product price information of the product to be sold, performing clustering operation on the product price information of the product to be sold by using K-means clustering algorithm to obtain clustering result, and setting a preset price level division rule according to the clustering result.

[0061] According to the preset product price level division rule, the product price level of the product to be sold is obtained.

[0062] In this embodiment, different prices of dishes are classified and divided into levels by using K-means clustering algorithm. Clustering is a process of dividing a data set into many groups. Data objects in the same group are similar, and data objects with large differences are in different groups. K-means clustering algorithm assumes that a data set contains n data objects, specifies the number of clusters k, and uses a division method to cluster objects in D into appropriate clusters, so that each object belongs to only one cluster. The algorithm specifies the value k as the number of clusters, places the data set to be clustered in a Euclidean space, randomly selects k centroids representing each cluster as the initial center point of the clustering process, and reassigns data objects to the nearest cluster according to the distance between each data object and each center point. The center point of the corresponding cluster is updated again after each iteration until there is no change. By using K-means clustering algorithm to perform clustering operation on the product price information of the product to be sold, the clustering result is obtained, and a preset price level division rule is set according to the clustering result. According to the preset product price level division rule, the product price level of the product to be sold is obtained.

[0063] Optionally, obtaining product price information of the product to be sold, performing clustering operation on the product price information of the product to be sold by using K-means clustering algorithm to obtain clustering result, and setting a preset price level division rule according to the clustering result, including:

[0064] According to the product price information of the product to be sold in each cluster in the clustering result, the mean price data corresponding to the product price information of the product to be sold in each cluster is calculated;

[0065] According to the range of mean price data in adjacent clusters, the level interval is divided to obtain the preset price level division rule.

[0066] In this embodiment, each cluster in the clustering result contains corresponding product prices. According to the product price of the product to be sold in each cluster in the clustering result, the mean price data corresponding to the product price information of the product to be sold in each cluster is calculated. According to the mean price in each cluster, the price level is divided. According to the range of mean price data in adjacent clusters, the level interval is divided to obtain the preset price level division rule.

[0067] It should be noted that when performing K-means clustering operation, the clustering validity index is used to determine whether the clustering result is reasonable. In this embodiment, the GA (generalization ability) is used to determine whether the clustering result is reasonable. The GA index is based on the current clustering result and evaluates the clustering result from the generalization ability in guided learning. That is, the merits of the clustering result are related to the generalization ability of the prediction of unknown samples. Therefore, it is different from the existing clustering validity index, whether it is an external validity index or an internal validity index. The GA index splits the obtained product price data to be sold into a training set and a data set, and performs clustering respectively. The clustering result of the training set is used for machine learning to construct a classifier, and the classifier is used to predict the test set, and then the prediction result and the clustering result are compared. According to the distance between the GA index value and 1, the rationality of clustering is determined. The closer the GA index value is to 1, the more reasonable the clustering result is.

[0068] In actual clustering, the number of clusters should not be too much, otherwise it will be difficult to explain the clustering result. Therefore, for limited optional product prices to be sold, an exhaustive method can be used. By calculating the GA index under different cluster numbers, the cluster number corresponding to the maximum GA index is selected as the most reasonable clustering result.

[0069] Optionally, according to the target consumption level of the export object and the product price level, a first quotation weight is determined, and an initial product quotation is generated according to the first quotation weight and the product price information, including:

[0070] According to the target consumption level of the export object and the preset score of each consumption level, a first ratio of the score of the target consumption level to the sum of the scores of each consumption level is calculated;

[0071] According to the product price level and the preset score of each price level, a second ratio of the score of the product price level to the sum of the scores of each price level is calculated;

[0072] The sum of the first ratio and the second ratio is calculated to obtain the first quotation weight, and the initial product quotation is generated according to the first quotation weight and the product price information.

[0073] In this embodiment, the consumption level and the product price level are combined and processed, and any consumption level and any price level are combined two by two. Different combinations set different values for the corresponding consumption level and price level, that is, the importance of each group to the quotation is set to different values. The consumption level and the product price level are combined to form a table, each table corresponds to a different array, and each array corresponds to different values of the corresponding consumption level and price level.

[0074] In the embodiment, different scores are set for each consumption level and each price level when the weights of different consumption levels and different price levels are set. For example, 5 points are set for the highest consumption level, 3 points are set for the middle consumption level, and 2 points are set for the low consumption level. 5 points are set for the highest price level, 3 points are set for the middle price level, and 2 points are set for the low price level. The first ratio of the score in each consumption level to the total score of each consumption level is calculated, and the second ratio of the score in each price level to the total score of each price level is calculated.

[0075] According to the target consumption level of the export target, the first ratio of the target consumption level to the total score of each consumption level is calculated. According to the product price level, the second ratio of the product price level to the total score of each price level is calculated. The sum of the first ratio and the second ratio is calculated to obtain the first quotation weight. According to the first quotation weight and the product price information, the initial product quotation is generated.

[0076] It should be noted that when the calculated initial product quotation is lower than the preset minimum initial product quotation or higher than the maximum preset product quotation, the initial product quotation is modified to the minimum initial product quotation or the maximum initial product quotation. For example, when the initial product quotation is lower than the minimum initial product quotation, the minimum initial product quotation is extracted as the initial product quotation of the corresponding product.

[0077] S204: Obtain trade portrait data of the export target.

[0078] In step S204, the trade portrait data of the export target is obtained. The trade portrait data includes the tariff information of the region where the export target is located, the currency tax rate information of the region where the export target is located and the local product to be sold, and the location information of the region where the export target is located.

[0079] In the embodiment, the trade portrait data of the export target is obtained using the crawler technology. When the relevant information page is obtained using the crawler technology, the web content is extracted to obtain the trade portrait data of the export target. During extraction, the structured and modular features of the web page are mainly used. The wrapper designs a unified template according to the layout rules of the web page, and obtains the location of the text in the page through analysis of the template. Currently, commonly used wrapper tools include TSIMMIS tool, XWRAR tool, etc. The TSIMMIS tool needs to manually write rules.

[0080] It should be noted that considering the characteristics of the webpage HTML document, the webpage body can also be extracted based on the method of establishing the document DOM tree. The basic idea of this method is to use the HTML mark of the webpage to represent the webpage as a document tree, to count the length of the text under each node of the document tree, the length of the link, the proportion of the number of text and link, and to determine whether the node is a body node.

[0081] S205: input the initial product quotation and trade portrait data into the document generation model, and output the export sales document for the export object.

[0082] In step S205, the document generation model is a model for generating an export sales document, and the generated export sales document includes the pre-quotation information of the product to be sold.

[0083] In this embodiment, the initial product quotation and trade portrait data are input into the document generation model, and the export sales document for the export object is output, wherein the document generation model is obtained according to the type of the consigned product, and different product types correspond to different document generation models.

[0084] It should be noted that when quoting multiple consigned products, the quotation task can be decomposed into multiple product quotation tasks, supporting simultaneous collaborative quotation of multiple products, and after each product is quoted, the quotation data of the project is automatically summarized, for example: for a large project including the quotation of two consigned products, A is responsible for the quotation of product A under the project, and B is responsible for the quotation of product B under the project, and the quotation of product A and product B is automatically summarized to generate the quotation data of the project, and displayed on the output export sales document.

[0085] Optionally, the initial product quotation and trade portrait data are input into the document generation model, and the export sales document for the export object is output, including:

[0086] Based on the initial product quotation and trade portrait data, an input matrix is constructed, and the input matrix is normalized to obtain target input data;

[0087] The target input data is input into the document generation model, and the export sales document for the export object is output.

[0088] In this embodiment, based on the initial product quotation and trade portrait data, an input matrix is constructed, and the input matrix is normalized to convert the initial product quotation and trade portrait data into data between 0 and 1, and obtain target input data. The normalized data can reduce the amount of calculation, and the target input data is input into the document generation model, and the export sales document for the export object is output.

[0089] The historical export object portrait data is processed by using a deep learning algorithm to mine data relationships, consumption portrait related factors of the historical export object are obtained, an evaluation model is constructed based on the consumption portrait related factors, the evaluation model is used to evaluate the consumption level of the export object, product price information of a product to be sold is obtained, a product price level of the product to be sold is determined by combining a preset price level division rule, a first offer weight is determined according to the target consumption level of the export object and the product price level, an initial product offer is generated according to the first offer weight and the product price information, trade portrait data of the export object is obtained, the trade portrait data includes tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and a local region of the product to be sold, and location information of the region where the export object is located, the initial product offer and the trade portrait data are input into a document generation model, and an export order for the export object is output, the corresponding product offer information is obtained according to the consumption level of the export object, the price level of the product to be sold by the user, and the corresponding regional factors, and the rationality of the product offer is increased.

[0090] Referring to Figure 3 is a flowchart of a portrait-based export order generation method provided by an embodiment of the present application, as Figure 3 The portrait-based export order generation method can include the following steps:

[0091] S301: The historical export object portrait data is processed by using a deep learning algorithm to mine data relationships, consumption portrait related factors of the historical export object are obtained, and an evaluation model is constructed based on the consumption portrait related factors;

[0092] S302: The consumption portrait data of the export object is obtained, the consumption level of the export object is evaluated by using the evaluation model based on the consumption portrait data, and the evaluation result is determined as the target consumption level of the export object;

[0093] S303: The product price information of the product to be sold is obtained, a product price level of the product to be sold is determined by combining a preset price level division rule, a first offer weight is determined according to the target consumption level of the export object and the product price level, and an initial product offer is generated according to the first offer weight and the product price information;

[0094] S304: The trade portrait data of the export object is obtained, the trade portrait data includes tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and a local region of the product to be sold, and location information of the region where the export object is located;

[0095] S305: The initial product offer and the trade portrait data are input into a document generation model, and an export order for the export object is output.

[0096] The steps S301 to S305 are the same as the contents of the steps S201 to S205, and the description of the steps S201 to S205 can be referred to, and details are not described herein again.

[0097] S306: Matching the export order with the preset sales statement to obtain a target sales statement, and sending the target sales statement to the corresponding client.

[0098] In this embodiment, in order to better sell the products, the corresponding sales statement is set for the export order, and when the product quotation in the export order is obtained through prediction, the corresponding sales statement is matched according to the quotation in the export order to obtain a target sales statement, and the target sales statement is sent to the corresponding client. When the salesperson receives the corresponding sales statement, the salesperson can better promote the products based on the sales statement.

[0099] Please refer to Figure 4 , Figure 4 is a structure diagram of an export order generation device based on a portrait provided by an embodiment of the present application. In this embodiment, each unit included in the mobile terminal is used to execute Figure 2 to Figure 3 each step in the corresponding embodiment. Please refer to Figure 2 to Figure 3 and Figure 2 to Figure 3 the related description in the corresponding embodiment. For ease of illustration, only the part related to this embodiment is shown. Please refer to Figure 4 , the generation device 40 includes a mining module 41, a consumption level evaluation module 42, an initial product quotation module 43, a trade portrait data acquisition module 44, and an export order generation module 45.

[0100] The mining module 41 is configured to perform data relationship mining processing on portrait data of historical export objects by using a deep learning algorithm, to obtain consumption portrait related factors of the historical export objects, and to construct an evaluation model based on the consumption portrait related factors. The level evaluation model is used to evaluate the consumption level of the export object.

[0101] The consumption level evaluation module 42 is configured to obtain consumption portrait data of the export object, to evaluate the consumption level of the export object based on the consumption portrait data by using the evaluation model, and to determine an evaluation result as a target consumption level of the export object.

[0102] The initial product quotation module 43 is configured to obtain product price information of a product to be sold, to determine a product price level of the product to be sold by combining a preset price level division rule, to determine a first quotation weight according to the target consumption level of the export object and the product price level, and to generate an initial product quotation according to the first quotation weight and the product price information.

[0103] The trade portrait data acquisition module 44 is configured to acquire trade portrait data of the export object, and the trade portrait data comprises tariff information of a region where the export object is located, currency tax rate information of a region where the export object is located and the product to be sold, and location information of the region where the export object is located.

[0104] The export order generation module 45 is configured to input the initial product quotation and the trade portrait data into a bill generation model, and output an export order for the export object.

[0105] Optionally, the mining module 41 comprises:

[0106] The consumption portrait related factor acquisition unit is configured to process portrait data of historical export objects by using an Apriori algorithm, and obtain consumption portrait related factors of the evaluation level according to the association between the consumption portrait related factors and the evaluation level in the portrait data of the historical export objects.

[0107] The construction unit is configured to analyze the consumption portrait related factors of the evaluation level, extract a feature engineering of the consumption portrait related factors of the evaluation level, and construct an evaluation model.

[0108] Optionally, the initial product quotation module 43 comprises:

[0109] The clustering unit is configured to acquire product price information of the product to be sold, perform clustering operation on the product price information of the product to be sold by using a K-means clustering algorithm, obtain a clustering result, and set a preset price level division rule according to the clustering result.

[0110] The price level division unit is configured to acquire the product price level of the product to be sold according to the preset product price level division rule.

[0111] Optionally, the clustering unit comprises:

[0112] The mean price acquisition subunit is configured to calculate mean price data corresponding to the product price information of the product to be sold in each cluster according to the product price information of the product to be sold in each cluster in the clustering result.

[0113] The rule division subunit is configured to divide a level interval according to a mean price data range in adjacent clusters, and obtain the preset price level division rule.

[0114] Optionally, the initial product quotation module 43 comprises:

[0115] The first ratio acquisition unit is configured to calculate a first ratio of a score of the target consumption level to a total score of scores of each consumption level according to the target consumption level of the export object and the preset score of each consumption level.

[0116] The second ratio acquisition unit is used to calculate a second ratio between the product price level score and the sum of the scores of each price level, based on the product price level and the preset score of each price level.

[0117] The initial product quotation unit is used to calculate the sum of the first ratio and the second ratio to obtain the first quotation weight, and to generate the initial product quotation based on the first quotation weight and the product price information.

[0118] Optionally, the aforementioned export order generation module 45 includes:

[0119] The target input data acquisition unit is used to construct an input matrix based on initial product quotations and trade profile data, and then normalize the input matrix to obtain the target input data.

[0120] The input unit is used to input the target input data into the document generation model and output the export order for the export target.

[0121] Optionally, the above-mentioned generating apparatus 40 further includes:

[0122] The sending module is used to match export orders with preset sales statements to obtain target sales statements, and then send the target sales statements to the corresponding clients.

[0123] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0124] Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Figure 5 As shown, the computer device of this embodiment includes: at least one processor ( Figure 5 Only one is shown in the diagram), a memory, and a computer program stored in the memory and capable of running on at least one processor, which, when executing the computer program, implements the steps in any of the above embodiments of the export order generation method based on the image.

[0125] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 5 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown in the illustration, or combinations of certain components, or different components, such as network interfaces, displays, and input devices.

[0126] The processor can be a CPU, and can also be other general-purpose processors, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0127] The memory includes a readable storage medium, an internal memory, etc., where the internal memory can be a memory of the computer device, and the internal memory provides an environment for running the operating system and the computer-readable instructions in the readable storage medium. The readable storage medium can be a hard disk of the computer device, and in other embodiments, can also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory can include both an internal storage unit of the computer device and an external storage device. The memory is used to store an operating system, an application program, a BootLoader, data, and other programs, such as program codes of computer programs, etc. The memory can also be used to temporarily store data that has been output or will be output.

[0128] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the unit and module in the above device can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here. If the integrated unit is realized in the form of 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 present application realizes all or part of the processes in the above-mentioned embodiment methods, which can be completed by a computer program to instruct related hardware. The computer program can be stored in a computer readable storage medium, and when the processor executes the computer program, the steps of the above-mentioned method embodiment can be realized. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form, etc. The computer readable medium can at least include any entity or device capable of carrying computer program code, recording medium, computer memory, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), electrical carrier signal, telecommunication signal and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0129] The present application realizes all or part of the processes in the above-mentioned embodiment methods, which can also be completed by a computer program product. When the computer program product runs on the computer device, it makes the computer device execute the steps that can realize the above-mentioned method embodiments.

[0130] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0131] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0132] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / computer device and method can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely schematic. The division of the modules or units is merely a logical function division, and there can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0133] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0134] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit it; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for generating export orders based on user profiles, characterized in that, The method for generating export orders includes: Deep learning algorithms are used to perform data relationship mining on the profile data of historical export targets to obtain consumption profile factors of the historical export targets, and an evaluation model is constructed based on the consumption profile factors; the evaluation model is used to evaluate the consumption level of export targets. Obtain consumer profile data of export targets, and use an evaluation model to evaluate the consumption level of export targets based on the consumer profile data, and determine the evaluation result as the target consumption level of export targets; Obtain product price information of the products to be sold, and determine the product price level of the products to be sold by combining the preset price level classification rules. Determine the first quotation weight according to the target consumption level of the export target and the product price level. Generate an initial product quotation according to the first quotation weight and the product price information. The step of determining a first pricing weight based on the target consumption level of the export target and the product price level, and generating an initial product price based on the first pricing weight and the product price information, includes: Based on the target consumption level of the export target and the preset score for each consumption level, a first ratio is calculated between the score of the target consumption level and the sum of the scores for each consumption level. Based on the product price level and the preset score for each price level, a second ratio is calculated between the product price level score and the sum of the scores for each price level. Calculate the sum of the first ratio and the second ratio to obtain the first price weight, and generate an initial product price based on the first price weight and the product price information; Obtain trade profile data of the export target, which includes tariff information of the region where the export target is located, the local currency tax rate information of the region where the export target is located and the local currency of the product to be sold, and the location information of the region where the export target is located. Input the initial product quotation and the trade profile data into the document generation model, and output the export order for the export target; The step of inputting the initial product quotation and the trade profile data into the document generation model and outputting an export order for the export target includes: Based on the initial product price and the trade profile data, an input matrix is ​​constructed, and the input matrix is ​​normalized to obtain the target input data; The target input data is input into the document generation model, and an export order for the export object is output.

2. The method for generating export orders based on profiles as described in claim 1, characterized in that, The step of using deep learning algorithms to perform data relationship mining on the profile data of historical export targets, obtaining consumption profile factors of the historical export targets, and constructing an evaluation model based on the consumption profile factors includes: The Apriori algorithm is used to process the profile data of historical export targets. Based on the correlation between consumption profile factors and evaluation levels in the profile data of historical export targets, the consumption profile factors of evaluation levels are obtained. The consumer profile factors of the evaluation level are analyzed, the feature engineering of the consumer profile factors of the evaluation level is extracted, and an evaluation model is constructed.

3. The method for generating export orders based on profiles as described in claim 1, characterized in that, The step of obtaining product price information of the products to be sold and determining the product price level of the products to be sold in combination with preset price level classification rules includes: Obtain product price information of products to be sold, perform clustering operation on the product price information of products to be sold using K-means clustering algorithm to obtain clustering results, and set preset price level division rules based on the clustering results; According to the preset product price level classification rules, the product price level of the product to be sold is obtained.

4. The method for generating export orders based on profiles as described in claim 3, characterized in that, The process involves obtaining product price information for products to be sold, performing clustering operations on the product price information using the K-means clustering algorithm to obtain clustering results, and setting preset price level classification rules based on the clustering results, including: Based on the product price information of the products to be sold in each cluster in the clustering results, the mean price data corresponding to the product price information of the products to be sold in each cluster is calculated. The price level classification rules are obtained by dividing the price range into grade intervals based on the mean price data range in adjacent clusters.

5. The method for generating export orders based on profiles as described in claim 1, characterized in that, After inputting the target input data into the document generation model and outputting the export order for the export target, the process further includes: The export order is matched with a preset sales statement to obtain the target sales statement, and the target sales statement is sent to the corresponding client.

6. A device for generating export orders based on portraits, characterized in that, The apparatus is used to implement the method of any one of claims 1-5, the apparatus comprising: The data mining module is used to perform data relationship mining on the profile data of historical export targets using deep learning algorithms, obtain the consumption profile factors of the historical export targets, and construct an evaluation model based on the consumption profile factors; the evaluation model is used to evaluate the consumption level of the export targets. The consumption level evaluation module is used to acquire consumption profile data of export targets, and based on the consumption profile data, use an evaluation model to evaluate the consumption level of the export targets, and determine the evaluation result as the target consumption level of the export targets. The initial product quotation module is used to obtain product price information of the products to be sold, and determine the product price level of the products to be sold by combining the preset price level classification rules. Based on the target consumption level of the export target and the product price level, a first quotation weight is determined, and an initial product quotation is generated based on the first quotation weight and the product price information. The trade profile data acquisition module is used to acquire trade profile data of the export target. The trade profile data includes tariff information of the region where the export target is located, the currency tax rate information of the region where the export target is located and the local currency of the product to be sold, and the location information of the region where the export target is located. The export order generation module is used to input the initial product quotation and the trade profile data into the document generation model and output the export order for the export target.

7. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, characterized in that, The computer-readable instructions are the export order generation method based on portraits as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the steps of the image-based export order generation method according to any one of claims 1-5.

Images