Shop operation optimization method and device, equipment and medium

Abstract

The invention relates to a shop operation optimization method and device, equipment and medium in the technical field of e-commerce, and the method comprises the steps: determining a plurality of shops of the same type, which belong to the same operation period and the same main operation category as a target shop, according to the operation period and the main operation category of the target shop; obtaining a plurality of operation values corresponding to the target shop and each shop of the same type, and determining an average operation index corresponding to each operation value based on the same operation values of all shops; when any operation value of the target shop is lower than the corresponding average operation index, at least one shop optimization program corresponding to the operation value is determined, and the shop optimization program is used for improving the operation value of the target shop; and constructing an application recommendation list containing the shop optimization program and pushing the application recommendation list to the target shop. The application can accurately identify the operation short board of the shop and recommend the program for making up the short board.

Description

Technical Field

[0001] This application relates to the field of e-commerce technology, and particularly to a method for optimizing store operations, as well as its corresponding device, computer device, and computer-readable storage medium. Background Art

[0002] In today's Internet era, the application markets on mobile and PC sides provide users with a rich variety of choices. Users can easily download and install various application programs to meet different needs. However, in the e-commerce scenario of independent websites, there are some problems in the current situation of the application market. The application market serving online stores operating on independent websites mainly provides merchants with various application programs related to the e-commerce industry, such as recommendation application programs for improving in-site conversion, cross-selling application programs, as well as SEO application programs and Push application programs for increasing store traffic. These application programs constitute an ecosystem aimed at helping merchants improve their operational effectiveness in multiple aspects of e-commerce store operations.

[0003] However, currently, the distribution efficiency of the application market for independent store is relatively low. Merchants often need to search for and install the required application programs by themselves. This passive approach results in most application programs lacking sufficient exposure and activity. After a new application program is launched, merchants may not be able to understand and pay attention to it in a timely manner, thus further reducing the possibility of application. In addition, the requirements for merchants are relatively high. They need to have a clear understanding of the application market, be able to accurately judge which application programs are useful to them, and select the most suitable one from among many similar application programs. This undoubtedly increases the decision-making cost for merchants.

[0004] Therefore, how to optimize the recommendation mechanism of the application market for independent store, reduce the decision-making cost for merchants, and increase the exposure and activity of application programs has become an urgent problem to be solved. Summary of the Invention

[0005] The primary objective of this application is to solve at least one of the above problems and provide a method for optimizing store operations, as well as its corresponding device, computer program product.

[0006] To meet the various objectives of this application, the following technical solutions are adopted in this application:

[0007] A method for optimizing store operations provided to meet one of the objectives of this application includes the following steps:

[0008] Determine multiple peer stores that belong to the same business cycle and the same main business category as the target store according to the business cycle and main business category of the target store;

[0009] Obtain multiple operation values corresponding to the target store and each of the similar stores respectively, and determine the average operation indicators corresponding to each operation value based on the same operation values of all stores;

[0010] When any one of the operation values of the target store is lower than its corresponding average operation indicator, determine at least one store optimization program corresponding to this operation value, and the store optimization program is used to increase this operation value of the target store;

[0011] Construct an application recommendation list containing the store optimization program and push it to the target store.

[0012] On the other hand, a store operation optimization device provided to meet one of the purposes of the present application includes a similarity determination module, an indicator determination module, a program determination module, and a list push module. Among them, the similarity determination module is used to determine multiple similar stores that belong to the same business cycle and the same main business category as the target store according to the business cycle and main business category of the target store; the indicator determination module is used to obtain multiple operation values corresponding to the target store and each of the similar stores respectively, and determine the average operation indicators corresponding to each operation value based on the same operation values of all stores; the program determination module is used to determine at least one store optimization program corresponding to this operation value when any one of the operation values of the target store is lower than its corresponding average operation indicator, and the store optimization program is used to increase this operation value of the target store; the list push module is used to construct an application recommendation list containing the store optimization program and push it to the target store.

[0013] On the other hand, a computer device provided to meet one of the purposes of the present application includes a central processing unit and a memory, and the central processing unit is used to call and run a computer program stored in the memory to execute the steps of the store operation optimization method described in the present application.

[0014] On the other hand, a computer program product provided to meet another purpose of the present application includes computer programs / instructions, and when the computer programs / instructions are executed by a processor, the steps of the method described in any embodiment of the present application are implemented.

[0015] The technical solution of the present application has many advantages, including but not limited to the following aspects:

[0016] This application screens similar stores that have the same operating cycle and main categories as the target store to ensure the relevance and reference value of subsequent comparative analysis and lay a reliable benchmark. Next, based on the multiple operational values corresponding to the target store and similar stores, the average indicator of each operational value is determined. When any operational value of the target store is lower than its corresponding average indicator, the store optimization program related to the indicator is analyzed, and a recommendation list is constructed based on the exposure effectiveness score of the program to ensure that the target store can obtain a store optimization program that matches the operational shortcomings in the current competitive market environment, thereby assisting store operations, extending operational shortcomings, and helping stores maintain competitiveness in a rapidly changing market environment.

[0017] In addition, this dynamic and real-time optimization method based on data not only reduces the cost of trial and error, but also significantly improves operational efficiency. It is especially suitable for stores with limited operating resources and time. It provides an efficient and sustainable solution for store operations. It can not only help target stores operate efficiently and make up for their shortcomings, but also provide them with long-term support in market competition. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0019] Figure 1 The network architecture of the e-commerce platform exemplified in this application;

[0020] Figure 2 This is a flowchart of a typical embodiment of the store operation optimization method of the present application;

[0021] Figure 3 This is a functional block diagram of the store operation optimization device of this application;

[0022] Figure 4 This is a schematic diagram of the structure of a computer device used in this application. DETAILED DESCRIPTION

[0023] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and are not to be construed as limiting the present application.

[0024] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the" and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of this application means the presence of the stated 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 their groups. It should be understood that when we say an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0025] Those skilled in the art can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless specifically defined as here.

[0026] As Figure 1 shown in the network architecture, the e-commerce platform 82 is deployed in the Internet to provide corresponding services to its users. Similarly, the devices 80 of the merchant users and the devices 81 of the consumer users of the e-commerce platform 82 are also connected to the Internet to use the services provided by the e-commerce platform.

[0027] An exemplary e-commerce platform 82 provides a supply-demand matching of products and / or services to the general public by means of the Internet infrastructure. In the e-commerce platform 82, the products and / or services are provided as commodity information. For the sake of simplicity of description, in this application, concepts such as commodities and products are used to refer to the products and / or services in the e-commerce platform 82, which may specifically be physical products, digital products, tickets, service subscriptions, other offline-fulfilled services, etc.

[0028] All parties in reality can access the e-commerce platform 82 in the name of users, use various online services provided by the e-commerce platform 82, and achieve the purpose of participating in the business activities realized by the e-commerce platform 82. These entities can be natural persons, legal persons, social organizations, etc. Corresponding to the two types of entities, merchants and consumers, in business activities, there are two major types of users on the e-commerce platform 82, namely merchant users and consumer users. All parties in the product circulation chain in business activities, including manufacturers, sellers, retailers, logistics providers, etc., can use online services on the e-commerce platform 82 in the identity of merchant users, while consumers in business activities, including actual or potential consumers, can use online services on the e-commerce platform 82 in the identity of their corresponding consumer users. In actual business activities, the same entity can act both as a merchant user and as a consumer user, and this should be understood flexibly.

[0029] The infrastructure for deploying the e-commerce platform 82 mainly includes a backend architecture and front-end devices. The backend architecture runs various online services through a service cluster, including middleware or front-end services for the platform side, services for consumers, services for merchants, etc., to enrich and improve its service functions; the front-end devices mainly cover the terminal devices used by users to access the e-commerce platform 82 as clients, including but not limited to various mobile terminals, personal computers, point-of-sale devices, etc. By way of example, merchant users can use their terminal device 80 to enter product information for their online stores, or generate their product information using the interfaces opened by the e-commerce platform; consumer users can access the web pages of the online stores realized by the e-commerce platform 82 through their terminal device 81, trigger the shopping process through the shopping buttons provided on the web pages, and call various online services provided by the e-commerce platform 82 during the shopping process, so as to achieve the purpose of placing an order for shopping.

[0030] In some embodiments, the e-commerce platform 82 can be implemented by a processing facility including a processor and a memory. The processing facility stores a set of instructions that, when executed, cause the e-commerce platform 82 to perform the e-commerce and support functions involved in this application. The processing facility can be part of a server, a client, a network infrastructure, a mobile computing platform, a cloud computing platform, a fixed computing platform, or other computing platforms, and provides the electronic components of the e-commerce platform 82, merchant devices, payment gateways, application developers, marketing channels, transportation providers, customer devices, point-of-sale devices, etc.

[0031] The e-commerce platform 82 can be implemented as online services such as cloud computing services, software as a service (SaaS), infrastructure as a service (IaaS), platform as a service (PaaS), desktop as a service (DaaS), hosted software as a service, mobile backend as a service (MBaaS), information technology management as a service (ITMaaS), etc. In some embodiments, the various functional components of the e-commerce platform 82 can be implemented to be suitable for operating on various platforms and operating systems. For example, for an online store, its administrator users enjoy the same or similar functions regardless of whether it is in various embodiments such as iOS, Android, HomonyOS, or the web.

[0032] The e-commerce platform 82 can implement corresponding independent sites for each merchant to run their corresponding online stores, and provide corresponding business management engine instances for the merchants to establish, maintain, and run one or more online stores in one or more independent sites. The business management engine instance can be used for content management, task automation, and data management of one or more online stores, and can configure various specific business processes of the online store through interfaces or built-in components, etc., to support the realization of business activities. The independent site is the infrastructure of the e-commerce platform 82 with cross-border service functions, and merchants can relatively centrally and independently maintain their online stores based on the independent site. The independent site usually has a domain name and storage space dedicated to the merchant, and there is relative independence between different independent sites. The e-commerce platform 82 can provide standardized or personalized technical support for a large number of independent sites, enabling merchant users to customize their own suitable business management engine instances and use this business management engine instance to maintain one or more online stores they own.

[0033] The online store can be configured and maintained in the background by the merchant user logging in to its business management engine instance as an administrator. With the support of various online services provided by the infrastructure of the e-commerce platform 82, the merchant user can, as an administrator, configure various functions in its online store and view various data. For example, the merchant user can manage all aspects of its online store, such as viewing the recent activities of the online store, updating the product catalog of the online store, managing orders, recent access activities, total order activities, etc.; the merchant user can also view more detailed information about the business and visitors to the merchant's online store by obtaining reports or metrics, such as showing the sales summary of the merchant's overall business, specific sales and participation data of the active sales and marketing channels, etc.

[0034] A store operation optimization method of the present application can be programmed as a computer program product and deployed to run in a client or a server. For example, in an exemplary application scenario of the present application, it can be deployed and implemented in the server of an e-commerce customer service platform. Thus, by accessing the interface opened after the computer program product runs, human-computer interaction can be performed with the process of the computer program product through a graphical user interface to execute this method.

[0035] Please refer to Figure 2 , in a typical embodiment of the store operation optimization method of the present application, the following steps are included:

[0036] Step S1100: Determine multiple peer stores that belong to the same business cycle and the same main business category as the target store according to the business cycle and the main business category of the target store;

[0037] Any store in the e-commerce platform can be used as the target store.

[0038] According to the business operation rules of stores in the e-commerce industry, the business process of online stores on an e-commerce platform can be divided into four business cycles, namely the probation period, the novice period, the development period, and the stable period. In order to divide the business cycles of each online store on the e-commerce platform, the basis for dividing the business cycles can include any one or more business indicators among the business duration range, transaction volume, visit volume range, product volume, consumer traffic range, consumption amount range, etc. Different business cycles can be divided based on the same or different business indicators. In one embodiment, stores with an operating days ≤ 14 days are divided into the probation period; stores with 14 days < operating days ≤ 30 days are divided into the novice period; stores with 30 days < operating days ≤ 90 days are divided into the development period; stores with operating days > 90 days are divided into the stable period. The left limit value and / or right limit value in each specific range can be set separately in advance by those skilled in the art as needed. In another embodiment, stores with an operating days ≤ 14 days, average daily sales ≤ 1000 yuan, average daily unique visitors (UV) ≤ 200, average daily page views (PV) ≤ 1000, and user repurchase rate ≤ 10% are divided into the probation period; stores with 14 days < operating days ≤ 30 days, 1000 yuan < average daily sales ≤ 5000 yuan, 200 < average daily unique visitors (UV) ≤ 500, 1000 < average daily page views (PV) ≤ 3000, and 10% < user repurchase rate ≤ 20% are divided into the novice period; stores with 30 days < operating days ≤ 90 days, 5000 yuan < average daily sales ≤ 10000 yuan, 500 < average daily unique visitors (UV) ≤ 1000, 3000 < average daily page views (PV) ≤ 8000, and 20% < user repurchase rate ≤ 30% are divided into the development period; stores with operating days > 90 days, average daily sales ≥ 10000 yuan, average daily unique visitors (UV) ≥ 1000, average daily page views (PV) ≥ 8000, and user repurchase rate ≥ 30% are divided into the stable period. The left limit value and / or right limit value in each specific range can be set separately in advance by those skilled in the art as needed.

[0039] Generally, the merchant users of each online store in the e-commerce platform provide the main business categories of their stores. In addition, in one embodiment, for each online store, the main product images and product titles of all the listed products in the store are obtained. The open-source pre-trained image encoding model and the open-source pre-trained text encoding model are used to encode the main product images and product titles of each product respectively, obtaining the corresponding image encoding vectors and text encoding vectors. The image encoding vectors and text encoding vectors are fused to obtain the text-image fusion vector corresponding to the product. The foregoing fusion can be implemented based on any one or more of weighted fusion, splicing fusion, Hadamard product fusion, PCA (Principal Component Analysis), mRMR (Maximal Relevance Minimal Redundancy Algorithm), multi-head attention mechanism, gated attention mechanism, and self-attention mechanism. Pre-training means that the corresponding encoding model has pre-learned the vectorized encoding representation of the input content. Further, a clustering algorithm is used to cluster the text-image fusion vectors of each product, determine each cluster, select the cluster with the most aggregated text-image fusion vectors, and input the vector representing the center of the cluster into the category division model that has been pre-trained to a converged state. The model infers a classification path in the category distribution system that best matches the vector, and takes this classification path as the main business category of the store. The vector representing the center of the cluster can be flexibly determined by those skilled in the art, and this step will not be elaborated here.

[0040] In the e-commerce field, the category distribution system is a hierarchical structure used to organize and manage category classifications. It is usually a tree-like classification framework used to divide a vast number of products into different category paths according to their attributes, uses, or characteristics. The category path belongs to a branch in the classification framework, and the lower the level in the branch, the finer the category division. Those skilled in the art can construct the category distribution system as needed based on the disclosure here and flexibly train the category division model, and this step will not be elaborated here. In the selection of the category division model, it can be an LSTM model, a Transformer model, etc., and those skilled in the art can select as needed.

[0041] First, obtain the business cycles and main business categories of each online store in the e-commerce platform determined above. Then, determine each other store whose business cycle and main business category are the same as those of the target store. These stores are the same type of stores that are in the same industry and the same period as the target store. Thus, they are used as the same-type stores respectively.

[0042] Step S1200: Obtain multiple operation values corresponding to the target store and each of the same-type stores respectively, and determine the average operation indicators corresponding to each operation value based on the same operation values of all stores.

[0043] Multiple operation values are scores corresponding to multiple operation dimensions, where the multiple operation dimensions include any combination of decoration design, store management, traffic attraction, conversion, marketing, after-sales, and logistics. The corresponding multiple operation values include any combination of decoration design score, store management score, traffic attraction score, conversion score, marketing score, after-sales score, and logistics score. For each store on the e-commerce platform, the various operation values of the store can be obtained through the following methods:

[0044] The decoration design score can be obtained by taking any one of the average page opening speed, average page stay duration, and page click-through rate of the store, or the value obtained by multiplying any combination of them by their respective weights and then summing them up. The weights of each item can be set by those skilled in the art as needed. The average page opening speed refers to the average value of the opening speeds of all pages in the store; the average page stay duration refers to the average value of the cumulative duration of all buyer users staying on the pages in the store; the page click-through rate refers to the value obtained by dividing the total number of pages clicked after being exposed in the store by the total number of exposed pages.

[0045] The store management score can be obtained by taking any one of the order processing duration, total quantity of recently listed products, and average customer service response speed of the store, or the value obtained by multiplying any combination of them by their respective weights and then summing them up. The weights of each item can be set by those skilled in the art as needed. The order processing duration refers to the average time from when a buyer pays for an order to when it is packed and shipped out of the store, calculated as the total sum of all order processing times divided by the total number of orders. The total quantity of recently listed products refers to the total number of new product SKUs in the store within a specified time period (such as the last 30 days). The average customer service response speed refers to the average value of the intervals between each user's consultation and the corresponding customer service response in all customer service conversations.

[0046] The traffic attraction score can be obtained by taking any one of the recent number of independent visitors (UV), total recent comments, total recent likes, total recent shares, total recent forwards, total recent add-to-cart quantity, and total recent orders of the store, or the value obtained by multiplying any combination of them by their respective weights and then summing them up. The weights of each item can be set by those skilled in the art as needed. The recent number of independent visitors (UV) refers to the total number of unique users who visited the store page within a specified time period (such as the last 30 days); the total recent comments / likes / shares / forwards refer to the cumulative number of product evaluations, likes, shares, and content forwards within a specified time period (such as the last 30 days) respectively; the total recent add-to-cart / order quantity: respectively count the number of times users add products to the cart and the total number of orders placed by users within a specified time period (such as the last 30 days).

[0047] The conversion score can be obtained by obtaining any one of the store's add-on rate, settlement rate, and conversion rate, or by multiplying any of these by their respective weights and adding the resulting value. The weights can be set as needed by those skilled in the art. The add-on rate refers to the ratio of users who add items to the total number of visiting users, calculated as follows: number of add-ons / number of page views × 100%; the settlement rate refers to the proportion of orders that are paid for after adding items to the store, that is, number of settled orders / number of add-on orders × 100%; and the conversion rate refers to the ratio of users who complete transactions to the total number of visiting users, calculated as follows: number of completed orders / number of page views × 100%.

[0048] The marketing score can be obtained by taking any one of the following: the total proportion of loyal customers of the store, the average repurchase period, or the sum of any number of factors multiplied by their respective weights. The weights can be set as needed by those skilled in the art. The proportion of loyal customers refers to the number of users who have repurchased at least twice within a specified time period (such as the past 30 days) as a percentage of the total number of customers; the average repurchase period refers to the average number of days between orders for all repeat customers.

[0049] The after-sales score can be obtained by obtaining any one of the store's average response time for after-sales returns, the ratio of negative reviews to the total, the return rate, and the average customer satisfaction score, or by multiplying any number of them by their respective weights and adding them together as the after-sales score. The weights can be set as needed by those skilled in the art. The average response time for after-sales returns refers to the average time it takes from the user submitting a return application to the merchant completing the processing; the ratio of negative reviews to the total refers to the ratio of the number of bad reviews (such as 1-2 star reviews) to the total number of all reviews; the return rate refers to the ratio of the number of return orders to the total number of completed orders; the average customer satisfaction score refers to the average of all users' scores on after-sales service.

[0050] The logistics score can be obtained by taking any one of the store's delivery on-time rate, package damage rate, and average shipping time, or by multiplying any number of these by their respective weights and adding them together. The weights can be set as needed by technicians in this field. The delivery on-time rate refers to the percentage of orders delivered on time, calculated as the number of on-time orders / total number of shipped orders × 100%; the package damage rate refers to the percentage of packages with damaged outer packaging or goods during transportation; the average shipping time refers to the average time from order payment to logistics collection, including the time spent on sorting, packaging, and other links.

[0051] In one embodiment, it can be understood that each operation value corresponds to an operation dimension. To quantitatively represent the average operation level of the same period of peers corresponding to this operation dimension, it is represented by the corresponding average operation index. Specifically in implementation, taking a single operation value as an example, and the same for other items. First, after summing up the operation values of the target store and each similar store corresponding to this item, divide by the total number of these stores to obtain the corresponding mean value as the basic mean value. Further, in order to more scientifically reflect the industry dynamics, avoid data deviation, and improve reliability and confidence, multiply the basic mean value by an adjustment parameter, and use the resulting product as the average operation index corresponding to this operation value.

[0052] Step S1300: When any operation value of the target store is lower than its corresponding average operation index, determine at least one store optimization program corresponding to this operation value, and the store optimization program is used to improve this operation value of the target store;

[0053] At this time, it indicates that the operation performance of the target store in the operation dimension corresponding to this operation value is not as good as the average level corresponding to similar stores in the same business cycle and main business category. Therefore, it is necessary to improve this operation value to make up for the deficiencies or shortcomings of the target store in this operation dimension. In one embodiment, according to the program category label associated with the operation dimension mapped to this operation value pre-configured, obtain the store optimization program belonging to this program category label.

[0054] It can be understood that an e-commerce platform can pre-deploy multiple store optimization programs and provide them for the stores in this platform to use, so as to assist the stores to operate better. The program category label to which the store optimization program belongs is used to represent the operation dimension applicable to the core function of this program. Different program category labels include any multiple of decoration design category, store management category, drainage category, conversion category, marketing category, after-sales category, logistics category. When each store optimization program is deployed, the program category label to which this program belongs can be marked manually or by a deep learning algorithm.

[0055] Step S1400: Construct an application recommendation list containing the store optimization program and push it to the target store.

[0056] Sort all store optimization programs in descending order according to the exposure effectiveness score, screen out multiple store optimization programs with higher rankings, and arrange these store optimization programs in the current ranking to form an application recommendation list and push it to the target store.

[0057] In one embodiment, for each store optimization program provided by the e-commerce platform for the stores therein, an exposure test is pre-conducted on the store optimization program. During the experiment, the corresponding click-through rate and installation rate are determined by listening to the program, and then the exposure effectiveness score is calculated based on the click-through rate and installation rate. The click-through rate refers to the value obtained by dividing the total number of times the store optimization program is clicked to view the program details after being exposed to the store by the total number of times exposed to the store; the installation rate refers to the value obtained by dividing the total number of times the store optimization program is installed after being exposed to the store by the total number of times exposed to the store.

[0058] An exemplary formula is as follows:

[0059]

[0060] Among them, Exposure score is the exposure effectiveness score, a and b correspond to the weights of the installation rate and the click-through rate respectively. Here, both weights can be set by those skilled in the art as needed. Install rate , Click rate correspond to the installation rate and the click-through rate respectively.

[0061] The exposure test refers to, for each of the above-provided store optimization programs, determining the business cycle that matches the program category label of the program based on a preset matching rule, screening out multiple stores belonging to the business cycle, exposing each store optimization program to the stores screened for it, and for each store optimization program, listening to whether the store optimization program is clicked by the store to view the program details after exposure and whether it is installed by the store, and accumulating the corresponding total number of exposures, total number of clicks, and total number of installations. Accordingly, the click-through rate and conversion rate corresponding to the program are calculated. The number of stores screened here can be set by those skilled in the art as needed.

[0062] The matching rule refers to, for each business cycle, establishing a matching relationship between the program category label to which the store optimization program applicable to the business dimension belongs and the business cycle according to the business dimension that needs to be most concerned about in the business cycle. For example: when the business cycle is the "probationary period", the business dimensions that need to be most concerned about are "decoration design, store management", and the program category labels matching the business cycle are "decoration design category, store management category"; when the business cycle is the "novice period", the business dimension that needs to be most concerned about is "traffic acquisition", and the program category label matching the business cycle is the "traffic acquisition category"; when the business cycle is the "development period", the business dimension that needs to be most concerned about is "conversion", and the program category label matching the business cycle is the "conversion category"; when the business cycle is the "stable period", the business dimensions that need to be most concerned about are "marketing, after-sales, logistics", and the program category labels matching the business cycle are "marketing category, after-sales category, logistics category".

[0063] It can be understood that after the application recommendation list is pushed to the target store, that is, the store optimization programs in the list are exposed to the target store. Therefore, each store optimization program in the list can be monitored to determine the new total exposure times, total click times, and total installation times, and accordingly, the exposure effectiveness scores of each store optimization program can be updated.

[0064] It is not difficult to understand from the above embodiments that compared with the prior art, the present application has multiple advantages, including at least:

[0065] The present application ensures the relevance and reference value of subsequent comparative analysis by screening similar stores that are consistent with the business cycle and main business categories of the target store, laying a reliable foundation. Immediately afterwards, based on multiple operation values corresponding to the target store and similar stores, the average indicators of each operation value are determined. When any operation value of the target store is lower than its corresponding average indicator, the store optimization programs related to this indicator are analyzed, and a recommendation list is constructed based on the exposure effectiveness scores of the programs, ensuring that the target store can obtain store optimization programs that match the operation shortboards existing in the current competitive market environment, thereby assisting store operation, lengthening the operation shortboards, and helping the store maintain competitiveness in the rapidly changing market environment.

[0066] In addition, such an optimization method based on data dynamically and in real time not only reduces the trial-and-error cost, but also significantly improves the operation efficiency. It is especially suitable for stores with limited operation resources and time, providing an efficient and sustainable solution for store operation. It can not only help the target store operate efficiently and make up for shortboards, but also provide long-term support for it in the market competition.

[0067] In a further embodiment, step S1200, determining the average operation indicators corresponding to each operation value based on the same operation values of all stores, includes the following steps:

[0068] Step S1210, for each operation value, add the operation values of all stores for this item to the corresponding data set, and calculate the first operation mean corresponding to all operation values in the data set;

[0069] After adding the operation values of all stores for this item to the data set initialized to be empty, divide the sum value obtained by adding up all the operation values in the data set by the total number of these operation values, and the obtained mean value is used as the first operation mean.

[0070] Step S1220, sort all the operation values in the data set, remove at least one operation value corresponding to the front and back in the sorted data set, and then calculate the second operation mean corresponding to the remaining operation means in the data set;

[0071] It can be to sort all the operation values in the dataset in descending or ascending order according to the operation values. Further, considering that avoiding the extreme operation values corresponding to the head and tail may lead to deviations inconsistent with the corresponding actual operation average level, at least one operation value corresponding to the front and back in the sorted dataset is removed, and the specific number of operation values to be removed can be set by those skilled in the art as needed. The sum value obtained by adding up all the remaining operation means in the dataset after removal is divided by the total number of these operation values, and the obtained mean value is used as the second operation mean value.

[0072] Step S1230: When the absolute value of the difference corresponding between the first operation mean value and the second operation mean value exceeds a preset threshold, use the second operation mean value as the average operation index corresponding to this item of operation data;

[0073] Calculate the absolute value of the difference obtained by subtracting the second operation mean value from the first operation mean value to obtain the absolute value of the difference. When the absolute value of the difference exceeds the preset threshold, this indicates that the deviation considered above has indeed occurred, that is, there is a certain deviation between the first operation mean value and the actual operation average level. Moreover, the second operation mean value should be closer to the actual operation average level than the first operation mean value. Therefore, use the second operation mean value as the average operation index corresponding to this item of operation data.

[0074] Step S1240: When the absolute value of the difference does not exceed the preset threshold, use the first operation mean value as the average operation index corresponding to this item of operation data.

[0075] This indicates that the deviation considered above has not occurred, that is, the deviation between the first operation mean value and the actual operation average level is within the tolerance range. Moreover, the first operation mean value should be closer to the actual operation average level than the second operation mean value. Therefore, use the first operation mean value as the average operation index corresponding to this item of operation data. Those skilled in the art can set the preset threshold as needed according to the above disclosure.

[0076] In this embodiment, first, the first operation mean value covers the operation values of all stores and comprehensively reflects the overall operation level. The second operation mean value is calculated after removing at least one operation value with a relatively high or low ranking in the dataset, which can effectively avoid the deviation caused by extreme values at the head and tail, making the result closer to the actual operation average level. By comparing the absolute difference between the first operation mean value and the second operation mean value, when the absolute difference exceeds the preset threshold, the second operation mean value is selected as the average operation index to ensure that the data after removing the influence of extreme values can more accurately represent the actual operation level. When the absolute difference does not exceed the preset threshold, the first operation mean value is used as the average operation index, making full use of all data information and avoiding information loss caused by excessive data removal. This flexible judgment mechanism takes into account both the integrity and accuracy of the data, enabling the finally determined average operation index to more truly reflect the actual operation status of the store, providing a more reliable basis for subsequent operation analysis, decision-making, etc. based on this index, helping to improve the store operation management level, optimize resource allocation, and enhance the market competitiveness of the store.

[0077] In a further embodiment, after step S1400 of constructing an application recommendation list including the store optimization program and pushing it to the target store, the following steps are included:

[0078] Step S1500: After monitoring that the target store has applied the store optimization program in the application recommendation list, generate a program execution log corresponding to the optimization effect information of the operation value improved by the store optimization program for this store.

[0079] After applying the store optimization program in the recommended list to the target store, the program execution log of the store optimization program is captured by calling the buried point interface or the log collection module. At the same time, the pre-encapsulated timed task module is called to apply the large language model to compare the operation values of the target store before and after using the program, so as to generate optimization effect information. The program execution log details the execution of the store optimization program in the target store, including the start time of the program, the execution duration from each start to end, the execution of specific function modules in each execution, and their input data and output results. For example, if the store optimization program is a tool for improving store decoration design, its program execution log can detail the following: First, record the specific time of each start of the tool, as well as the total duration from start to end and the execution duration of each function module. Second, record which function modules are used, such as page layout adjustment, color scheme selection, picture material upload, etc. For the input data, the page layout adjustment module will record the parameters of the original page layout, such as module position, size, etc.; the color scheme selection module will record the original color scheme, such as background color, text color, etc.; the picture material upload module will record the name, format, size, etc. of the uploaded pictures. In terms of output results, the page layout adjustment module will record the adjusted page layout parameters, such as the new module position, size, etc.; the color scheme selection module will record the new color scheme, such as the new background color, text color, etc.; the picture material upload module will record the position, size, etc. of the uploaded pictures in the page.

[0080] The optimization effect information quantitatively presents the improvement of the operation values of the target store after applying the store optimization program. Taking the above decoration design optimization program as an example, the optimization effect information reflects that after applying the program, the average page opening speed has increased by 10%, the average page stay duration has increased by 20%, the page click-through rate is 5%, and the overall decoration design score has increased by 4%.

[0081] The large language model is suitable for text processing in the NLP field. It is pre-trained on an extremely large corpus until convergence to acquire the ability to generate human language, and has a certain degree of accurate text semantic understanding ability and logical reasoning ability. The model selection includes Falcon, Chinchilla, PaLM, LLaMA 2, text-embedding-ada-002, etc.

[0082] Step S1510, when the optimization effect information meets the expected optimization conditions, generate an optimization compliance report based on the program execution log and the optimization effect information, and push it to the target store;

[0083] The expected optimization conditions are pre-set criteria for measuring whether the store optimization program has effectively improved the target store operation value using a preset threshold, and those skilled in the art can set the preset threshold as needed. Taking the optimization program for improving the store decoration design score as an example, if the expected optimization condition is to increase the decoration design score of the target store by 2%, then when the optimization effect information indicates that the decoration design score of the target store has actually increased by 2% or more, it is considered that the expected optimization condition has been met.

[0084] By applying the large language model to synthesize the program execution log and optimization effect information, with the goal of comprehensively demonstrating the application effect of the store optimization program in the target store, an optimization compliance report is generated. The report content includes the specific optimization strategies applied and the specific optimization measures implemented for the specific components of the store, as well as a process description corresponding to the entire process from the start of optimization to reaching compliance. For example, for the optimization program to increase the conversion rate, the report will detail how the optimization program optimizes the product detail page of the target store, such as adjusting the page layout and optimizing the product recommendation algorithm, and how these optimization measures improve the conversion-related indicators such as the add-to-cart rate and settlement rate of the target store, ultimately achieving the compliance of the conversion score, so that the target store can clearly understand the effectiveness of the optimization program and its positive impact on store operations.

[0085] In this embodiment, by comprehensively monitoring the operation situation and effect of the optimization program in the target store, when the corresponding optimization effect reaches the standard, a detailed optimization compliance report is generated and pushed to the store, providing the store with a clear summary of the optimization results and in-depth analysis of the optimization details, thereby enhancing the store's trust and satisfaction with the optimization program.

[0086] In a further embodiment, in response to the program usage feedback event triggered by the target store, the conclusion information in the optimization compliance report is encapsulated as comment information of the store optimization program and pushed to the preset comment interface of the program.

[0087] When the operation staff of the target store (usually the merchant user) approves the content of the optimization compliance report after reviewing it and believes that the store optimization program has indeed helped the store operation, they can trigger a program usage feedback event through the terminal. Then, the server of the e-commerce platform responds to this event, extracts key conclusions from the optimization compliance report, such as the conclusive statement "The store optimization program has effectively improved the conversion score of the target store by 4%", and encapsulates it into comment information that conforms to the preset format. The preset format can be set according to the requirements of the evaluation interface.

[0088] The preset comment interface is an interface specifically set by the store optimization program for collecting user feedback, which can be flexibly implemented by those skilled in the art. After the server pushes the encapsulated comment information to this interface, the comment information will be parsed and displayed in the user evaluation area of the store optimization program for other stores to refer to, helping other stores better understand the actual effect and applicability of this optimization program, so as to make a decision on whether to select this program. The display method can be in the form of a vertical scrolling carousel, a form with text and icons, etc., which can be selected and implemented by those skilled in the art.

[0089] In this embodiment, by encapsulating the key conclusions in the optimization compliance report as user comment information and pushing it to the preset comment interface, the active collection and display of application feedback are realized without the need for the target store to edit comments, providing a true and reliable reference basis for other stores, promoting the popularization and application of the optimization program, and driving the healthy development of the entire store optimization ecosystem.

[0090] In a further embodiment, after step S1500, monitoring the program execution log generated after the target store applies the store optimization program in the application recommendation list and the optimization effect information corresponding to the operation value improvement of the target store by this store optimization program, the following steps are included:

[0091] Step S1501, when the optimization effect information does not meet the expected optimization conditions, obtain the store portrait of the target store and obtain a set of successful cases. The set of successful cases includes store portraits, optimization effect information, and program execution logs corresponding to multiple optimized and compliant stores. The optimization effect information and program execution logs are obtained by monitoring the application of the store optimization program to the corresponding optimized and compliant stores, and each optimization effect information meets the expected optimization conditions, and the main business categories and business cycles in the store portraits corresponding to each optimized and compliant store and the target store are the same;

[0092] It can be understood that whenever the optimization effect information corresponding to each target store in the e-commerce platform meets the expected optimization conditions after applying the store optimization program, each target store is regarded as an optimized qualified store, and the store portrait of each optimized qualified store is obtained. The optimization effect, program execution log, and store portrait corresponding to each optimized qualified store are associated with the program unique identifier of the store optimization program to form corresponding successful cases, which are stored in the database for future reference. For each store in the e-commerce platform, a store portrait of the store can be constructed. The store portrait is a set of store features formed through multi-dimensional data modeling, including the business cycle, main business category, various operation values, and customer group distribution of the corresponding store. The customer group distribution includes loyal customers, ordinary customers, customers prone to churn, lost customers, and the specific quantity corresponding to each type of customer. For the specific classification of each type of customer, a deep learning algorithm can be used to model the association between customer categories and multiple customer behavior dimensions, so that according to the specific values corresponding to each customer behavior dimension of the customer, the customer category to which the customer belongs can be determined, and those skilled in the art can implement it flexibly. Each customer behavior dimension includes any multiple of the frequency of visiting the store, average page stay duration, purchase frequency, amount per purchase, and repurchase frequency. The program unique identifier is unique and is used to refer to the corresponding single store optimization program to distinguish it from other store optimization programs. For example, ID, and those skilled in the art can set it flexibly.

[0093] At this time, the unqualified processing process is triggered and executed. After calling the data acquisition interface to obtain the store portrait of the target store, multiple successful cases in the database are retrieved and obtained. The store portrait of the optimized qualified store in each of these successful cases includes the main business category and business cycle in the store portrait of the target store, and the store optimization program is the store optimization program applied to the target store. These successful cases are aggregated to form a successful case set.

[0094] Step S1502: Compare and analyze the differences between the store portrait, program execution log, and optimization effect information corresponding to the target store and the store portrait, program execution log, and optimization effect information corresponding to the optimized qualified store in the successful case set, and generate an optimization unqualified report, which is pushed to the target store.

[0095] The store portrait, program execution log, and optimization effect information of the target store and the successful case set are used as context information to be embedded in a preset prompt text template to obtain the corresponding prompt text. The prompt text is input into the large language model to obtain the optimization unqualified report output by the model.

[0096] The prompt text template includes a task description and context information to be embedded. The task description can be: "You are a senior e-commerce operation optimization analyst. Please find the store portrait, program execution logs, and optimization effect information of the target store, and the obvious differences between the store portrait, program execution logs, and optimization effect information corresponding to the optimized and qualified stores in the successful case set. Summarize each difference item and output it as an optimization non-compliance report." To ensure the quality of the generated optimization non-compliance report, the large language model can be pre-finely tuned until the training converges, and then applied to the report generation in this step. Those skilled in the art can flexibly implement the fine-tuning training of the large language model.

[0097] In this embodiment, when the optimization effect of applying the store optimization program in the target store fails to meet the standard, by comparing and analyzing the target store with multiple successful cases to generate a corresponding report and pushing it to the target store, it can alleviate the dissatisfaction of the target store and help the target store summarize operation experience based on the report and operate better in the future.

[0098] In a further embodiment, after step S1502, comparing and analyzing the differences between the store portrait, program execution logs, and optimization effect information corresponding to the target store and the store portrait, program execution logs, and optimization effect information corresponding to the optimized and qualified stores in the successful case set to generate an optimization non-compliance report and pushing it to the target store, the following steps are included:

[0099] Step S2500: Collect the optimization non-compliance reports corresponding to each store, cluster each optimization non-compliance report, and determine multiple clusters;

[0100] After generating the optimization non-compliance report, collect the optimization non-compliance reports corresponding to all stores, and classify the reports with similar characteristics or commonalities into the same cluster through a clustering algorithm. Specifically, natural language processing techniques based on text similarity (such as TF-IDF combined with cosine similarity calculation) or numerical clustering methods based on multi-dimensional feature vectors (such as K-means, hierarchical clustering, etc.) can be used.

[0101] Step S2510: For each cluster, conduct attribution analysis based on each optimization non-compliance report belonging to the cluster to obtain a corresponding set of optimization non-compliance factors;

[0102] For each cluster, perform attribution analysis to extract common optimization non-compliance factors, and collect these factors to form an optimization non-compliance factor set. This process conducts multi-dimensional correlation analysis by comparing the operational value differences among stores within the cluster, abnormal points in program execution logs, and customer portrait characteristics. For example, all stores in a certain cluster have a "conversion score lower than the average index". By analyzing the program logs using a large language model, it is found that the product association recommendation module is generally not enabled when these stores apply the conversion optimization program, and the customer portrait shows that 30% of their visitors are price-sensitive users. At this time, the core factor can be attributed to "insufficient matching degree between product recommendation strategy and user portrait". The attribution method can use a decision tree model trained by artificial feature engineering and / or a large language model to identify optimization non-compliance factors, or verify the influence weight of specific factors on the optimization result through A / B testing, and then determine whether it belongs to an optimization non-compliance factor accordingly.

[0103] Step S2511: Generate a corresponding program application report based on the optimization non-compliance factor set of each cluster and push it to the preset application feedback interface of the store optimization program.

[0104] For each cluster, use a large language model to summarize the main reasons for the optimization non-compliance of the cluster based on the optimization non-compliance factor set of the cluster. Then, on this basis, the model sorts out the main reasons for the optimization non-compliance of each category to form a program application report for output. Specifically, when implementing, the large language model can be guided to complete the above output in the way of chain of thought and few-shot examples.

[0105] The preset application feedback interface is an interface specifically set by the store optimization program for collecting feedback after the program is put into online application, and can be flexibly implemented by those skilled in the art.

[0106] In this embodiment, through clustering and attribution analysis of multiple collected optimization non-compliance reports, and then forming reports on the non-compliance reasons of multiple categories obtained and pushing them to the application feedback interface, it can clearly provide the deficiencies of the current optimization operation for the store optimization program, contribute to the iteration of the program, and provide better optimization operation services.

[0107] In a further embodiment, step S1400: Construct an application recommendation list including the store optimization program and push it to the target store, including the following steps:

[0108] Step S1410: Obtain the store portrait of the target store and the optimization strategy text corresponding to each store optimization program;

[0109] It can be understood that usually, program R & D engineers will optimize the program for each store and edit the corresponding description text. The large language model can be used to determine the description content of how the program optimizes the store's operation from the description text of the store optimization program, generate an optimization strategy text, and output it.

[0110] Step S1420: For each store optimization program, use a preset effectiveness evaluation model to evaluate the exposure effectiveness score of the program according to the store portrait and the optimization strategy text of the program.

[0111] The effectiveness evaluation model has been pre-trained to a convergent state and has learned the ability to evaluate the corresponding exposure effectiveness score according to the input store portrait and optimization strategy text.

[0112] For the training of the effectiveness evaluation model, a training set is prepared in advance, which includes multiple training samples and their supervision labels. The training sample is the store portrait of the relative store and the optimization strategy text of the store optimization program applied to the store. The supervision label of the training sample is the actual exposure effectiveness score based on the change value of the operation value of the store after applying the program.

[0113] The model structure of the effectiveness evaluation model includes a text representation layer and a classifier connected later. The text representation layer can be a deep learning model suitable for text semantic representation in the NLP field, such as RoBERTa, Transfomer Encoder, MPNet, BiLSTM, GPT, etc.; the classifier can be a fully connected layer or an MLP (multi-layer perceptron).

[0114] Call a single training sample and its supervision label in the training set, input them into the effectiveness evaluation model. The text representation layer in the effectiveness evaluation model extracts the deep semantic information of the input training sample, embeds and represents it as a corresponding text feature vector. Then, the classifier maps the text feature vector to the classification space and outputs the predicted exposure effectiveness score. Call a preset loss function to calculate the loss value corresponding to the predicted exposure effectiveness score according to the supervision label of the training sample. When the loss value reaches the preset threshold, it indicates that the effectiveness evaluation model has been trained to a convergent state, and thus the training of the effectiveness evaluation model can be terminated; when the loss value does not reach the preset threshold, it indicates that the effectiveness evaluation model has not converged. Therefore, the model is updated by gradient according to the loss value, usually by backpropagation to correct the weight parameters of each link of the model to make the model further approach convergence. Then, continue to call other training samples and their supervision labels to perform iterative training on the effectiveness evaluation model until the model is trained to a convergent state. The loss function and the preset threshold can be set by those skilled in the art as needed.

[0115] Step S1430: Construct an application recommendation list according to the exposure effectiveness scores of each store optimization program and push it to the target store.

[0116] Sort all store optimization programs in descending order according to the exposure effect score, select multiple store optimization programs with higher rankings, and arrange these store optimization programs in the current order to form an application recommendation list and push it to the target store.

[0117] In this embodiment, by using the effect evaluation model to evaluate the exposure effect score of the program based on the store portrait of the store and the optimization strategy text of the store optimization program, it is ensured that the score has accuracy, reliability and interpretability, making the constructed recommendation list more convenient for the store to select reliable programs. In addition, it can solve the cold start problem of determining the exposure effect of the unexposed store optimization program when it is first recommended to the store.

[0118] In a further embodiment, after step S1430, constructing an application recommendation list and pushing it to the target store, the following steps are included:

[0119] Step S1431: Monitor each store optimization program in the application recommendation list, and determine the usage frequency, usage duration, and change value of the operation value corresponding to each program applied to each store;

[0120] The running data of each store optimization program in each store can be captured in real time through the built-in log collection module, including the number of program starts, that is, the usage frequency, the duration of each single run, that is, the usage duration, and the specific value corresponding to the increase or decrease of the operation value of the store before and after the program runs, or the specific value corresponding to remaining unchanged is 0, that is, the change value. For example, a certain drainage program is started 5 times (usage frequency) in store A, runs for an average of 30 minutes each time (usage duration), and the decoration design score increases by 10% (change value).

[0121] Step S1432: Update the exposure effect score corresponding to each store optimization program based on the usage frequency, usage duration, and change value corresponding to each store.

[0122] For each store optimization program, multiply the usage frequency, usage duration, and change value corresponding to its application to each store by their respective weights and then sum them up to obtain the latest exposure effect score, and replace the current exposure effect score of the store optimization program with the latest exposure effect score. Each weight can be set by those skilled in the art as needed.

[0123] In this embodiment, by updating the exposure effect score of the program based on the usage frequency, usage duration, and change value corresponding to the application of the store optimization program to each store, the real-time and reliability of the score are ensured.

[0124] Please refer to Figure 3, A store operation optimization device provided to meet one of the purposes of this application is a functional embodiment of the store operation optimization method of this application. On the other hand, a store operation optimization device provided to meet one of the purposes of this application includes a similar store determination module 1100, an index determination module 1200, a program determination module 1300, and a list push module 1400. Among them, the similar store determination module 1100 is used to determine multiple similar stores that belong to the same business cycle and the same main business category as the target store according to the business cycle and main business category of the target store; the index determination module 1200 is used to respectively obtain multiple operation values corresponding to the target store and each of the similar stores, and determine the average operation index corresponding to each operation value based on the same-item operation values of all stores; the program determination module 1300 is used to determine at least one store optimization program corresponding to the operation value when any operation value of the target store is lower than its corresponding average operation index, and the store optimization program is used to improve the operation value of the target store in this item; the list push module 1400 is used to construct an application recommendation list containing the store optimization program and push it to the target store.

[0125] In a further embodiment, the index determination module 1200 includes: a first mean calculation sub-module, which is used for each operation value, adding the operation values of all stores in this item to the corresponding data set, and calculating the first operation mean corresponding to all operation values in the data set; a second mean calculation sub-module, which is used to sort all operation values in the data set, remove at least one operation value corresponding to the front and back in the data set sorting, and then calculate the second operation mean corresponding to the remaining operation means in the data set; a first index determination sub-module, which is used when the absolute value of the difference corresponding to the first operation mean and the second operation mean exceeds a preset threshold, taking the second operation mean as the average operation index corresponding to this operation data; a second index determination sub-module, which is used when the absolute value of the difference does not exceed the preset threshold, taking the first operation mean as the average operation index corresponding to this item of operation data.

[0126] In a further embodiment, after the list push module 1400, it includes: a first program monitoring sub-module, configured to monitor the program execution log generated after the target store applies the store optimization program in the application recommendation list and the optimization effect information corresponding to the operation value improvement of the target store by the store optimization program; a first report push sub-module, configured to generate an optimization compliance report based on the program execution log and the optimization effect information and push it to the target store when the optimization effect information meets the expected optimization conditions; a first interface push sub-module, configured to, in response to a program usage feedback event triggered by the target store, encapsulate the conclusion information in the optimization compliance report into the comment information of the store optimization program and push it to the preset comment interface of the program.

[0127] In a further embodiment, after the program monitoring sub-module, it includes: a first data acquisition sub-module, configured to, when the optimization effect information does not meet the expected optimization conditions, acquire the store portrait of the target store and acquire a successful case set, where the successful case set includes the store portraits, optimization effect information, and program execution logs corresponding to multiple optimized and compliant stores, the optimization effect information and the program execution logs are obtained by monitoring the store optimization program applied to the corresponding optimized and compliant stores, and each piece of optimization effect information meets the expected optimization conditions, and the main business categories and business cycles in the store portraits corresponding to each optimized and compliant store and the target store are the same; a second report push sub-module, configured to perform a difference comparison analysis on the store portrait, program execution log, and optimization effect information corresponding to the target store and the store portrait, program execution log, and optimization effect information corresponding to the optimized and compliant stores in the successful case set, and generate an optimization non-compliance report and push it to the target store.

[0128] In a further embodiment, after the second report push sub-module, it includes: a report clustering sub-module, configured to collect the optimization non-compliance reports corresponding to each store, cluster the optimization non-compliance reports, and determine multiple clusters; an attribution analysis sub-module, configured to, for each cluster, perform an attribution analysis based on the optimization non-compliance reports belonging to the cluster to obtain a corresponding set of optimization non-compliance factors; a second interface push sub-module, configured to generate a corresponding program application report based on the sets of optimization non-compliance factors of each cluster and push it to the preset application feedback interface of the store optimization program.

[0129] In a further embodiment, before the list pushing module 1400, it includes: a second data acquisition sub-module, configured to acquire the store portrait of the target store and the optimization strategy text corresponding to each store optimization program; a scoring estimation sub-module, configured to, for each store optimization program, use a preset effectiveness evaluation model to evaluate the exposure effectiveness score of this program according to the store portrait and the optimization strategy text of this program; a list pushing sub-module, configured to construct an application recommendation list according to the exposure effectiveness scores of each store optimization program and push it to the target store.

[0130] In a further embodiment, after the list pushing sub-module, it includes: a second program monitoring sub-module, configured to monitor each store optimization program in the application recommendation list and determine the usage frequency, usage duration, and change value of the operation value corresponding to each program applied to each store; a scoring update sub-module, configured to update the exposure effectiveness scores corresponding to each store optimization program based on the usage frequency, usage duration, and change value corresponding to each store.

[0131] To solve the above technical problems, an embodiment of the present application also provides a computer device. As Figure 4 shown, it is a schematic internal structure diagram of the computer device. The computer device includes a processor, a computer-readable storage medium, a memory, and a network interface connected through a system bus. Among them, the computer-readable storage medium of the computer device stores an operating system, a database, and computer-readable instructions. The control information sequence can be stored in the database. When the computer-readable instructions are executed by the processor, the processor can implement a store operation optimization method. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. Computer-readable instructions can be stored in the memory of the computer device. When the computer-readable instructions are executed by the processor, the processor can execute the store operation optimization method of the present application. The network interface of the computer device is used to connect and communicate with the terminal. Those skilled in the art can understand that Figure 4 the structure shown in

[0132] In this embodiment, the processor is used to execute Figure 3The specific functions of each module and its sub-modules therein, the memory stores program codes and various types of data required to execute the above-mentioned modules or sub-modules. The network interface is used for data transmission between user terminals or servers. The memory in this embodiment stores the program codes and data required to execute all modules / sub-modules in the store operation optimization device of the present application, and the server can call the program codes and data of the server to execute the functions of all sub-modules.

[0133] The present application also provides a storage medium storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors are caused to execute the steps of the store operation optimization method according to any embodiment of the present application.

[0134] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments of the present application can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above methods. Among them, the aforementioned storage medium can be a computer-readable storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0135] In summary, the present application can accurately identify the operation shortboards of stores and recommend programs to make up for the shortboards.

[0136] Those skilled in the art of the present technology can understand that the steps, measures, and solutions in the various operations, methods, and processes discussed in the present application can be alternated, changed, combined, or deleted. Further, the other steps, measures, and solutions in the various operations, methods, and processes discussed in the present application can also be alternated, changed, rearranged, decomposed, combined, or deleted. Further, the steps, measures, and solutions in the prior art that are open source in the present application can also be alternated, changed, rearranged, decomposed, combined, or deleted.

[0137] The above are only some embodiments of the present application. It should be noted that for those of ordinary skill in the art of the present technology, without departing from the principle of the present application, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A method for optimizing store operations, characterized in that, It includes the following steps: According to the business cycle and main business category of the target store, determine multiple peer stores that belong to the same business cycle and the same main business category as the target store; Respectively obtain multiple operation values corresponding to the target store and each of the peer stores, and based on the same-item operation values of all stores, determine the average operation indicators corresponding to each operation value; When any one of the operation values of the target store is lower than its corresponding average operation indicator, determine at least one store optimization program corresponding to this operation value, and the store optimization program is used to improve this operation value of the target store; Construct an application recommendation list containing the store optimization program and push it to the target store.

2. The store operation optimization method according to claim 1, wherein Based on the same-item operation values of all stores, determining the average operation indicators corresponding to each operation value includes the following steps: For each operation value, add the operation values of all stores for this item to the corresponding data set, and calculate the first operation average value corresponding to all operation values in the data set; Sort all the operation values in the data set, remove at least one operation value corresponding to the top and bottom in the data set sorting, and then calculate the second operation average value corresponding to the remaining operation average values in the data set; When the absolute value of the difference between the first operation average value and the second operation average value exceeds the preset threshold, use the second operation average value as the average operation indicator corresponding to this operation data; When the absolute value of the difference does not exceed the preset threshold, use the first operation average value as the average operation indicator corresponding to this operation data.

3. The store operation optimization method according to claim 1, wherein, After constructing an application recommendation list containing the store optimization program and pushing it to the target store, it includes the following steps: Monitor the program execution log generated after the target store applies the store optimization program in the application recommendation list and the optimization effect information corresponding to the improvement of the operation value of the target store by this store optimization program; When the optimization effect information meets the expected optimization conditions, generate an optimization compliance report based on the program execution log and the optimization effect information, and push it to the target store.

4. The store operation optimization method according to claim 3, wherein, After monitoring the program execution log generated after the target store applies the store optimization program in the application recommendation list and the optimization effect information corresponding to the improvement of the operation value of the target store by this store optimization program, it includes the following steps: When the optimization effect information does not meet the expected optimization conditions, obtain the store portrait of the target store and obtain a successful case set. The successful case set includes the store portraits, optimization effect information, and program execution logs corresponding to multiple optimization-compliant stores. The optimization effect information and program execution logs are obtained by monitoring the store optimization programs applied to the corresponding optimization-compliant stores, and each optimization effect information meets the expected optimization conditions, and the main business category and business cycle in the store portraits corresponding to each optimization-compliant store and the target store are the same; Perform a differential comparison analysis on the store portrait, program execution log, and optimization effect information corresponding to the target store with those corresponding to the optimized compliant stores in the success case set, generate an optimization non-compliance report, and push it to the target store.

5. The store operation optimization method according to claim 4, wherein After performing a differential comparison analysis on the store portrait, program execution log, and optimization effect information corresponding to the target store with those corresponding to the optimized compliant stores in the success case set, generating an optimization non-compliance report, and pushing it to the target store, the following steps are included: Collect the optimization non-compliance reports corresponding to each store, cluster each optimization non-compliance report, and determine multiple clusters. For each cluster, perform a root cause analysis based on the optimization non-compliance reports belonging to that cluster to obtain the corresponding set of optimization non-compliance factors. Generate a corresponding program application report based on the sets of optimization non-compliance factors for each cluster and push it to the preset application feedback interface of the store optimization program.

6. The store operation optimization method according to claim 1, wherein, Construct an application recommendation list containing the store optimization program and push it to the target store, including the following steps: Obtain the store portrait of the target store and the optimization strategy texts corresponding to each store optimization program. For each store optimization program, use a preset effectiveness evaluation model to evaluate the exposure effectiveness score of the program based on the store portrait and the optimization strategy text of the program. Construct an application recommendation list based on the exposure effectiveness scores of each store optimization program and push it to the target store.

7. The store operation optimization method according to claim 6, characterized in that After constructing an application recommendation list and pushing it to the target store, the following steps are included: Monitor each store optimization program in the application recommendation list, and determine the usage frequency, usage duration, and change values of the operation values corresponding to each program when applied to each store. Update the exposure effectiveness scores corresponding to each store optimization program based on the usage frequency, usage duration, and change values corresponding to each store.

8. An apparatus for optimizing store operations, characterized in that, Include: A similar store determination module for determining multiple similar stores that belong to the same business cycle and the same main business category as the target store based on the business cycle and main business category of the target store. An index determination module for respectively obtaining multiple operation values corresponding to the target store and each similar store, and determining the average operation index corresponding to each operation value based on the same operation values of all stores. A program determination module for determining at least one store optimization program corresponding to an operation value when any operation value of the target store is lower than its corresponding average operation index, where the store optimization program is used to improve the operation value of the target store for that item. A list push module for constructing an application recommendation list containing the store optimization program and pushing it to the target store.

9. A computer device, comprising a central processing unit and a memory, characterized in that, The central processing unit is used to call and run the computer program stored in the memory to execute the steps of the method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program implemented by the method according to any one of claims 1 to 7 in the form of computer-readable instructions. When the computer program is called and run by a computer, it executes the steps included in the corresponding method.

Images