Recommended methods, devices, computer equipment, and storage media for safe deposit box categories
By acquiring data on item categories and climate characteristics, and using a trained safe deposit box category recommendation model, the problem of inaccurate safe deposit box category recommendations was solved, resulting in more accurate recommendations, reducing the risk of item damage, and improving customer experience.
Patent Information
- Application Number
- CN202310125310.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-02-07
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-02-07
AI Technical Summary
In existing technologies, the method of recommending safe deposit box categories cannot accurately take into account the category of the items and climate differences, resulting in inaccurate recommendations and potential damage to the items.
By acquiring data on the categories of items to be stored and the climate characteristics of the storage areas, a trained safe deposit box category recommendation model is used to determine the recommendation probability of each safe deposit box category. Based on the probability, the most suitable safe deposit box category is recommended. The model training process includes data preprocessing, gradient descent algorithm, and dynamic learning rate optimization.
It enables the recommendation of appropriate safe deposit box categories based on differences in item type and climate characteristics, reducing the risk of item damage and improving the accuracy of recommendations and customer experience.
Smart Images

Figure CN116070864B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, computer equipment, storage medium, and computer program product for recommending safe deposit box categories. Background Technology
[0002] As people's living standards and incomes continue to improve, the demand for safe deposit boxes or vaults is gradually increasing. Currently, users who need to store physical items must apply for vault services online. The recommended vault type is usually chosen by the user themselves, or based on recommendations from service personnel.
[0003] However, due to differences in item categories and variations in the products of different manufacturers and versions of safe deposit boxes, the computer sometimes relies solely on previously used safe deposit box category data when recommending storage box categories. Alternatively, it may passively respond to user or customer actions to match and recommend safe deposit box categories. Therefore, these methods fail to provide accurate safe deposit box category recommendations. Summary of the Invention
[0004] Therefore, it is necessary to provide an accurate method, apparatus, computer device, computer-readable storage medium, and computer program product for recommending safe deposit box categories, in response to the above-mentioned technical problems.
[0005] Firstly, this application provides a method for recommending safe deposit box categories. The method includes:
[0006] Obtain the category and storage location of the items to be stored;
[0007] Query the climate characteristics data of the stored region;
[0008] Based on the category of the items to be stored and the climate characteristics of the storage area, determine the recommended probability of the safe category;
[0009] Based on the recommendation probability of each safe deposit box category, determine the safe deposit box categories to be recommended.
[0010] In one embodiment, determining the recommended probability of a safe deposit box category based on the category of the items to be stored and the climate characteristics of the storage area includes:
[0011] Using the categories of items to be stored and the climate characteristics of the storage area as input data, the trained safe deposit box category recommendation model is called to obtain the recommendation probability of the safe deposit box category;
[0012] The safe deposit box category recommendation model was trained based on climate characteristic data of the storage area and box type characteristic data.
[0013] In one embodiment, before using the category of the items to be stored and the climate characteristics of the storage area as input data, and calling a trained safe deposit box category recommendation model to obtain the recommendation probability of the safe deposit box category, the method further includes:
[0014] Obtain model sample data, which includes climate distribution data and corresponding box feature data for different regions;
[0015] Data preprocessing is performed on climate distribution data and corresponding box-shaped feature data from different regions to obtain sample data for the target model.
[0016] Obtain the initial safe deposit box category recommendation model;
[0017] Based on the target model sample data, an initial safe deposit box category recommendation model is trained to obtain the safe deposit box category recommendation model.
[0018] In one embodiment, data preprocessing is performed on climate distribution data and box-shaped feature data from different regions to obtain target model sample data, including:
[0019] The box type feature data are sequentially processed by numericalization and clustering to obtain the classification results of the safe deposit box categories.
[0020] Based on climate distribution data from different regions, different regions are divided to obtain climate characteristic data for each region;
[0021] The results of the classification of safe deposit boxes and the corresponding climate characteristic data of different regions are collected to obtain the target model sample data.
[0022] In one embodiment, based on the target model sample data, an initial safe deposit box category recommendation model is trained to obtain the safe deposit box category recommendation model, which includes:
[0023] Based on the target model sample data, the initial safe deposit box category recommendation model is trained using the gradient descent algorithm and dynamic learning rate, thus obtaining the safe deposit box category recommendation model.
[0024] In one embodiment, the safe deposit box categories to be recommended are determined based on the recommendation probability of each category, including:
[0025] Rank the recommendation probabilities of safe deposit box categories;
[0026] Select a first preset number of safe deposit box categories with a high probability of recommendation;
[0027] The selected safe deposit box category is determined to be the recommended safe deposit box category.
[0028] In one embodiment, after determining the type of safe deposit box to be recommended, the method further includes:
[0029] Push notifications for recommended safe deposit box categories;
[0030] Receive feedback data on item storage evaluation;
[0031] The safe deposit box category recommendation model is updated based on the feedback data on item storage evaluation.
[0032] In one embodiment, after pushing the recommended safe deposit box category, the method further includes:
[0033] Obtain the business location corresponding to the safe deposit box category to be recommended, as well as the location of the recommended safe deposit box category;
[0034] The push notification recommends a second preset number of business locations that are close to the location of the target object in the safe deposit box category.
[0035] Secondly, this application also provides a safe deposit box category recommendation device. The device includes:
[0036] The data acquisition module is used to obtain the category and storage location of the items to be stored;
[0037] The data query module is used to obtain climate characteristic data of the stored area;
[0038] The recommendation probability determination module is used to determine the recommendation probability of the safe category based on the category of the item to be stored and the climate characteristic data of the storage area;
[0039] The safe deposit box category determination module is used to determine the safe deposit box category to be recommended based on the recommendation probability of the safe deposit box category.
[0040] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps in the aforementioned recommended method for safe deposit box categories.
[0041] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the steps in the aforementioned method for recommending safe deposit box categories.
[0042] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the steps in the aforementioned recommended method for safe deposit box categories.
[0043] The aforementioned method, apparatus, computer equipment, storage medium, and computer program products for recommending safe deposit box categories use the category of the items to be stored and the storage location as benchmarks. They also consider the impact of climate differences in different regions on the storage requirements, further querying climate characteristic data for the storage location. Based on the category of the items to be stored and the climate characteristics of the storage location, the recommendation probability of a safe deposit box category is determined, and then the recommended safe deposit box category is determined based on the recommendation probability. This entire solution not only considers the different requirements for the type of box for the stored items due to different categories, but also takes into account the potential damage caused by climate differences. By comprehensively considering the category of the items to be stored and the climate characteristics of the storage location, it provides targeted recommendations for safe deposit box categories that are more suitable for the items to be stored, achieving accurate safe deposit box category recommendations. Furthermore, it can assist business personnel or guide customers in determining the final, more suitable safe deposit box category and effectively reduce the risk of damage to the stored items due to inaccurate safe deposit box category recommendations. Attached Figure Description
[0044] Figure 1 This is a diagram illustrating the application environment of a safe deposit box category recommendation method in one embodiment.
[0045] Figure 2 This is a flowchart illustrating a safe deposit box category recommendation method in one embodiment;
[0046] Figure 3 This is a flowchart illustrating the safe deposit box category recommendation method in another embodiment;
[0047] Figure 4 This is a flowchart illustrating the process of training a safe deposit box category recommendation model in one embodiment;
[0048] Figure 5 This is a schematic diagram of the structure of a safe deposit box category recommendation model in one embodiment;
[0049] Figure 6 This is a schematic diagram of neuron operations in a safe deposit box category recommendation model in one embodiment;
[0050] Figure 7 This is a detailed flowchart illustrating the safe deposit box category recommendation method in another embodiment;
[0051] Figure 8 This is a structural block diagram of a safe deposit box category recommendation device in one embodiment;
[0052] Figure 9 This is a structural block diagram of a safe deposit box category recommendation device in another embodiment;
[0053] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation
[0054] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0055] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0056] The safe deposit box category recommendation method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed in the cloud or on another network server. Specifically, an object needing to store items can log into terminal 102, operate through the interactive interface of terminal 102, input the category and storage location of the items to be stored, and send a safe deposit box category recommendation request message carrying the category and location of the items to be stored to server 104. Server 104 receives the safe deposit box category recommendation request message, obtains the category and location of the items to be stored, then queries the climate characteristic data of the storage location, and determines the recommendation probability of the safe deposit box category based on the category and climate characteristic data of the storage location. Finally, based on the recommendation probability, it determines the safe deposit box category to be recommended. Furthermore, server 104 can return the recommended safe deposit box category to terminal 102. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle systems. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.
[0057] In one embodiment, such as Figure 2 As shown, a method for recommending safe deposit box categories is provided, which can be applied to... Figure 1 Taking server 104 as an example, the following steps are included:
[0058] Step S202: Obtain the category and storage area of the items to be stored.
[0059] The categories of items to be stored can include, but are not limited to, documents, paintings, precision instruments, and contracts. The storage area refers to the region where the items will be stored. In this embodiment, considering the characteristics of climate distribution, the storage area can be divided by province or city. In this embodiment, the category of items to be stored and the storage area can be input by the user who needs to store the items. In another embodiment, the storage area can be a pre-set area or the location of a recommended storage box category.
[0060] Step S204: Query the climate characteristic data of the stored region.
[0061] Climate characteristic data includes climate type, corresponding climate characteristics such as temperature, humidity, and air pH, as well as whether centralized heating is available. Climate types include subtropical climate, tropical rainforest climate, tropical savanna climate, tropical desert climate, and temperate oceanic climate. Due to significant climate differences between regions, such as the vast differences in temperature and humidity between coastal and inland provinces, the storage conditions for items also vary considerably. That is, the suitable storage location for the same item will change depending on the climate. For example, documents such as contracts are not suitable for storage in humid and rainy areas, as they are prone to dampness and mold, leading to varying degrees of damage. In practical applications, developers can pre-divide regions according to their climate distribution and establish a correspondence between each region and climate characteristic data. In this embodiment, after obtaining the storage location of the item to be stored, the climate characteristic data of the storage location can be queried based on the storage location and the correspondence between the region and the climate characteristic data. In other embodiments, a third-party climate characteristic data query service can be invoked to query the climate characteristic data of the storage location.
[0062] Step S206: Determine the recommended probability of the safe deposit box category based on the category of the items to be stored and the climate characteristics data of the storage area.
[0063] In this embodiment, "safe deposit box" can be understood as a general term for safe deposit boxes, safes, or vaults. Specifically, safe deposit boxes can be categorized into various types based on differences in dimensions such as size, material, shock resistance, fire resistance, and moisture resistance. It should be noted that safe deposit boxes with very similar sizes and materials can be classified into the same category. In this embodiment, based on the existing types of safe deposit boxes and combined with dimensions such as size, material, shock resistance, fire resistance, and moisture resistance, safe deposit boxes can be divided into 46 categories.
[0064] Following the above embodiments, after obtaining the category of the items to be stored and the climate characteristics data of the storage area, the category of the items to be stored and the climate characteristics data of the storage area can be used as input data to call a classifier and obtain the recommendation probability of each locker category. In other embodiments, a locker type prediction model or a locker category recommendation model can be built and trained based on deep learning or machine learning algorithms, and then the category of the items to be stored and the climate characteristics data of the storage area can be input into the trained locker type prediction model or locker category recommendation model to obtain the recommendation probability of each locker category.
[0065] Step S208: Determine the safe deposit box category to be recommended based on the recommendation probability of the safe deposit box category.
[0066] The types of safe deposit boxes to be recommended can be multiple or a single. Continuing from the previous embodiment, after obtaining the recommendation probability corresponding to each safe deposit box type, the box type with the highest recommendation probability can be selected as the safe deposit box category to be recommended. In another embodiment, the top few box types with relatively high recommendation probabilities can also be selected as the safe deposit box categories to be recommended. Furthermore, in another embodiment, the safe deposit box categories to be recommended can be pushed to the terminal to assist users and business personnel in selecting the safe deposit box category that best matches the items to be stored.
[0067] The aforementioned method for recommending safe deposit box categories uses the category of the items to be stored and the storage location as benchmarks. It also considers the impact of climate differences in different regions on the storage requirements, further querying climate characteristic data for the storage location. Based on the category of the items to be stored and the climate characteristics of the storage location, the recommendation probability of a safe deposit box category is determined, and then the recommended safe deposit box category is determined based on the recommendation probability. This entire solution not only considers the different requirements for the type of box due to the different categories of items, but also takes into account the impact of climate differences on the storage of items, which may lead to damage. By comprehensively considering the category of the items to be stored and the climate characteristics of the storage location, it provides targeted recommendations for safe deposit box categories that are more suitable for the items to be stored, achieving accurate recommendations. Furthermore, it can assist business personnel or guide customers in determining the final, more suitable safe deposit box category and effectively reduce the risk of damage to the stored items due to inaccurate safe deposit box category recommendations.
[0068] like Figure 3 As shown, in one embodiment, step S206 includes: step S226, taking the category of the item to be stored and the climate characteristic data of the storage area as input data, calling the trained safe deposit box category recommendation model to obtain the recommendation probability of the safe deposit box category.
[0069] In this embodiment, developers can collect a large amount of box-type feature data and climate feature data of the storage area. Using the climate feature data of the storage area and the box-type feature data as model sample data, a box-type recommendation model is trained. The trained box-type recommendation model can output the recommendation probability for each box type category. Specifically, the category of the item to be stored and the climate feature data of the storage area can be used as input data and fed into the trained box-type recommendation model. The trained box-type recommendation model is then invoked, enabling it to predict the output probability of each box type category and obtain the recommendation probability for that box type. In this embodiment, by invoking the trained box-type recommendation model, the recommendation probability of the box type category can be obtained quickly, accurately, and objectively based on the category of the item to be stored and the climate feature data of the storage area.
[0070] like Figure 4 As shown, in one embodiment, before using the category of the item to be stored and the climate characteristics data of the storage area as input data, and calling a trained safe deposit box category recommendation model to obtain the recommendation probability of the safe deposit box category, the method further includes:
[0071] Step S205: Obtain model sample data, which includes climate distribution data and corresponding box feature data for different regions.
[0072] Step S225: Perform data preprocessing on climate distribution data and corresponding box-shaped feature data for different regions to obtain target model sample data.
[0073] Step S245: Obtain the initial safe deposit box category recommendation model.
[0074] Step S265: Based on the target model sample data, train the initial safe deposit box category recommendation model to obtain the safe deposit box category recommendation model.
[0075] Target model sample data refers to the model sample data obtained after preprocessing the model sample data. Considering the global nature of the safe deposit box business, climate distribution data can refer to the spatial distribution data of global climate types. In this embodiment, climate distribution data can be a climate distribution map. Box type characteristic data includes the safe deposit box's size, material, shock resistance, moisture resistance, fire resistance, and lock type (such as combination lock, mechanical lock, fingerprint lock), etc.
[0076] Following the above embodiments, before calling the safe deposit box category recommendation model, the model can be trained as follows: First, obtain climate distribution data and box type feature data for different regions. Then, perform data preprocessing on the climate distribution data and corresponding box type feature data for different regions to obtain target model sample data. Data preprocessing may include data cleaning, data quantification, and data transformation. Then, based on the complexity of the preprocessed data, design an initial safe deposit box category recommendation model containing a 6-layer neural network, including an input layer, hidden layers, and an output layer. For the specific model structure, please refer to [link to relevant documentation]. Figure 5 , where x i For input data, y i This outputs the recommendation probabilities for each safe deposit box category. Specifically, the input layer consists of 36-dimensional feature data. There are three hidden layers: the first layer has 60×128 nodes (7680 nodes); the second layer has 128×256 nodes (32768 nodes); the third layer has 256×512 nodes (131072 nodes); and the fourth layer has 512×1 nodes (512 nodes). The output layer displays the recommendation probabilities for 46 safe deposit box categories. After constructing the initial safe deposit box category recommendation model, the preprocessed target model sample data can be split proportionally: 70% training data, 20% test data, and 10% validation data. The training data is used for model training, the test data is used for model testing, and the validation data is used to verify the model's prediction accuracy. The order of each data point is then randomly shuffled.
[0077] The specific model training process can be as follows: Based on the training data, determine the dimensionality of the model input data. After obtaining the dimensionality of the input data, begin model training. In this embodiment, the hidden layer principle is as follows:
[0078]
[0079] Where h is the output of the neuron, f is the activation function of the neuron, and x i It is the input of the i-th other neuron, where n is the total number of inputs, w i is the weight parameter of the i-th other neuron, and b is the threshold of the current neuron. The neuron computation graph can be found in [reference needed]. Figure 6 .
[0080] The activation function used in this invention is the ReLU activation function, and its principle is as follows:
[0081]
[0082] Where x is the input parameter.
[0083] Since the output layer provides recommendation probabilities for 46 different safe deposit box categories, and cross-entropy is used to characterize the distance between two probability distributions, cross-entropy can be used as the loss function in this embodiment.
[0084] First, the output of the hidden layer network needs to be calculated using the sigmoid algorithm to output probability values:
[0085]
[0086] Where z is the network output value, j is the output probability value of the j-th safe deposit box category, K is the number of safe deposit box categories, k is the k-th safe deposit box category in the denominator summation operation, and σ(z) j It is the value obtained after performing a sigmoid calculation on the output value.
[0087] Then, cross-entropy loss is calculated between it and the real sample:
[0088]
[0089] Where i represents the i-th type of safe deposit box, n represents the number of safe deposit box types, and in this embodiment, there are 46 types of safe deposit boxes. i () represents the probability of an item being stored in the current actual safe deposit box category. The matrix representation of the actual samples for a single box category is p(x) = [0,1], where 0 indicates no item is stored, 1 indicates item is stored, and q represents the probability of an item being stored in the safe deposit box category. i () represents the probability that the current safe deposit box category is predicted to be 1 (to be stored). The probability of the predicted output sample matrix being 1 is:
[0090] q i ()=P(p i ()=1|)
[0091] Meanwhile, model training can employ a gradient descent algorithm. In this embodiment, the Adam gradient descent algorithm can be used, and a dynamic learning rate can be used for the model. The principle of the Adam gradient descent algorithm is as follows:
[0092]
[0093] Where t is the number of iterations, initialized to 0; w t For the parameter to be updated, w t-1 The parameters are as follows: α is the dynamic learning rate; For m t Correction, For v t The correction ε is a constant added to maintain numerical stability; in this embodiment, ε = 10. -8 .
[0094] and The formula is:
[0095]
[0096]
[0097] Wherein, β1 and β2 are constants, representing the exponential decay rates of the first and second moments, respectively, controlling the exponential decay. In this embodiment, β1 = 0.9 and β2 = 0.999. and The powers of t are β1 and β2; m t For the first moment estimate of the gradient, v t This is the second moment estimate of the gradient.
[0098] m t With v t The formula is:
[0099] m t =1×m t-1 +(1-1)×g t
[0100]
[0101] Among them, g t It is the gradient at time step t.
[0102] The principle of dynamic learning rate is as follows:
[0103]
[0104] in, The learning rate is α, the input learning rate is ε, the decay rate is ε, and step is the total number of training steps for the model. In this embodiment, the decay rate is 0.9, and the number of training iterations is 10,000. In practice, the training data can be divided into multiple batches. If the number of training iterations is less than 200, the next batch of data is selected sequentially to train the network parameters of the initial safe deposit box category recommendation model. If the number of training iterations reaches 200, test data is used to test the error data until the maximum number of training iterations is reached, completing the model training and saving the trained safe deposit box category recommendation model. The trained safe deposit box category recommendation model is then validated using validation data, simulating a real-world safe deposit box type recommendation scenario. The validated initial safe deposit box category recommendation model is the final trained safe deposit box category recommendation model. It is understood that the gradient descent algorithm can also be stochastic gradient descent, full gradient descent, or other gradient descent algorithms; this is not limited here.
[0105] In this embodiment, the safe deposit box category recommendation model is trained using the Adam gradient descent algorithm and a dynamic learning rate, which ensures more stable parameter updates and makes the model more stable.
[0106] In one embodiment, step S225 includes: step S245, performing numerical processing and clustering processing on the box-type feature data in sequence to obtain the box category classification results, dividing different regions according to the climate distribution data of different regions to obtain the climate feature data of different regions, and collecting the box category classification results and the corresponding climate feature data of different regions to obtain the target model sample data.
[0107] In this embodiment, the process of preprocessing the model sample data can be as follows: The formula for setting the safe deposit box category is:
[0108] P i =luster(T i )
[0109] Among them, T i This represents the numericalized model sample data. Cluster is a clustering algorithm for similar category safe deposit boxes. In this embodiment, based on expert experience, the safe deposit boxes are numerically quantified and clustered according to dimensions such as material, size, thickness, shock resistance, moisture resistance, fire resistance, and lock type (e.g., combination lock, mechanical lock, fingerprint lock). i This is the result of clustering, specifically the classification of safe deposit boxes. In detail, the safe deposit boxes can be classified into 46 categories, represented numerically as [1, 46].
[0110] Clustering first requires numerical operations on dimensions such as material and size in the box-shaped feature data:
[0111] T i =umerical(S i )
[0112] Where S i Let be the sample of the corresponding dimension, i be the sample of the i-th dimension, and Numerical be the numerical value of the sample of the corresponding dimension, such as defining a sample that supports shock resistance as 1 and a sample that does not support shock resistance as 0. The corresponding clustering formula is:
[0113]
[0114] Where, λ iIn this embodiment, the weighting coefficients for factors such as the size of the storage box, shock resistance, and fire resistance are all hard screening criteria. For example, if the dimensions (length, width, and height) of the items to be stored exceed 30×30×30, a storage box that cannot hold the items will not be recommended. Provided the storage box size meets the customer's needs, the size is given a higher weight in the weighting and calculation. However, after screening for shock resistance and fire resistance, if the storage area does not have storage boxes with shock resistance and fire resistance functions, but the customer urgently needs to store fragile items, then a storage box without shock resistance and fire resistance will be recommended. However, the customer will be informed that the storage box does not have these functions, and they will be asked to make their own decision.
[0115] Because the safe deposit box business is global, in this embodiment, after obtaining climate distribution data for different regions, the regions can be divided into 12 climate types based on their climate, including: subtropical climate, tropical rainforest climate, tropical savanna climate, tropical desert climate, and temperate oceanic climate. Simultaneously, regions are also divided based on whether or not they have centralized heating to avoid discrepancies caused by human factors within the same climate zone.
[0116] C i =Climate(L j )
[0117] Among them, C i The corresponding climate codes are used, such as 0 for tropical rainforest climate and 1 for tropical savanna climate. "Climate" represents the climate classification method. In this embodiment, climate classification can be performed based on a climate distribution map, and some provinces and cities are selected for experimental classification at the city level. Additionally, 0 and 1 are used to indicate whether centralized heating is provided.
[0118] After preprocessing the container type feature data and climate feature data from different regions to obtain the container category classification results and corresponding climate feature data for different regions, the container category classification results and corresponding climate feature data for different regions can be aggregated to obtain the target model sample data. In this embodiment, by sequentially performing numerical processing and clustering processing on the container type feature data, and encoding the climate distribution data of different regions, the corresponding regional climate feature data is obtained, enabling the data to be successfully used for model training.
[0119] like Figure 3 As shown, in one embodiment, step S208 includes: step S228, sorting the recommendation probabilities of safe deposit box categories, selecting a first preset number of safe deposit box categories with higher recommendation probabilities, and obtaining safe deposit box categories to be recommended.
[0120] In specific implementation, the determination of the safe deposit box category to be recommended can be as follows: The recommendation probabilities of the previously obtained safe deposit box categories are sorted, and then a first preset number of safe deposit box categories with higher recommendation probabilities are selected. In this embodiment, the top three safe deposit box categories with higher recommendation probabilities can be selected as the categories to be recommended. In other embodiments, the determination of the categories to be recommended can also be based on the top three safe deposit box categories with higher recommendation probabilities output by the trained safe deposit box category recommendation model. It is understood that the sorting method can be either sequential or descending, and is not limited here. In other embodiments, the top five safe deposit box categories with higher recommendation probabilities can also be selected as the categories to be recommended. The number of categories to be recommended can be determined according to the actual situation and is not limited here. In this embodiment, by selecting a first preset number of safe deposit box categories with higher recommendation probabilities as the categories to be recommended, accurate recommendations of safe deposit box categories can be achieved, enabling customers or business personnel to quickly determine more suitable safe deposit boxes.
[0121] like Figure 7 As shown, in one embodiment, after step S208, the method further includes: step S210, pushing the recommended safe deposit box category, receiving the item storage evaluation feedback data, and updating the safe deposit box category recommendation model based on the item storage evaluation feedback data.
[0122] The item storage evaluation feedback data includes the final determined safe deposit box category and evaluation data on item storage provided by sales personnel or customers. In this embodiment, the item storage evaluation feedback data can be the final determined safe deposit box category and item storage satisfaction data. In practical applications, after determining the safe deposit box category to be recommended, the category can be pushed to the client so that customers or sales personnel can select the final safe deposit box category. Furthermore, customers or sales personnel can input the final determined safe deposit box category and their satisfaction rating for this safe deposit box category recommendation. The server can further update the safe deposit box category recommendation model based on the received item storage evaluation feedback data, thereby optimizing the safe deposit box category recommendation model. In this embodiment, updating the safe deposit box category recommendation model based on the item storage evaluation feedback data enables further optimization of the safe deposit box category recommendation model, achieving more accurate safe deposit box category recommendations.
[0123] like Figure 7 As shown, in one embodiment, after pushing the recommended safe deposit box category, the method further includes: step S212, obtaining the business location corresponding to the recommended safe deposit box category and the location of the recommended safe deposit box category, and pushing a second preset number of business locations that are close to the location of the recommended safe deposit box category.
[0124] Business location can refer to a physical business location. The safe deposit box category recommendation target refers to the entity that initiates the safe deposit box category recommendation request, typically a customer. In practical applications, after pushing the recommended safe deposit box category, authorization can be obtained by acquiring the customer's geographical location (i.e., the customer's location). Simultaneously, the recommended business locations with safe deposit boxes corresponding to the requested category are queried. Then, the distance between the customer's location and the recommended business locations is calculated, and the five closest business locations are selected and pushed to the client. It is understood that the number of business locations pushed can be determined based on actual circumstances and is not limited here. In this embodiment, by pushing the business locations closest to the customer's location, the likelihood of the customer traveling to a distant business location can be reduced, saving customer time and improving customer experience.
[0125] To provide a clearer explanation of the recommended methods for safe deposit box categories provided in this application, the following is in conjunction with the appendix. Figure 7 and one A specific embodiment will be described below, which includes the following steps:
[0126] Step S202: Obtain the category and storage area of the items to be stored.
[0127] Step S204: Query the climate characteristic data of the stored region.
[0128] Step S226: Using the category of the items to be stored and the climate characteristics of the storage area as input data, call the trained safe deposit box category recommendation model to obtain the recommendation probability of the safe deposit box category.
[0129] The safe deposit box category recommendation model can be obtained in the following way: obtain model sample data, which includes climate distribution data and corresponding box type feature data of different regions; perform data preprocessing on the climate distribution data and corresponding box type feature data of different regions to obtain target model sample data; obtain the initial safe deposit box category recommendation model; and train the initial safe deposit box category recommendation model based on the target model sample data using the Adam gradient descent algorithm to obtain the safe deposit box category recommendation model.
[0130] Step S228: Sort the recommendation probabilities of safe deposit box categories, select a first preset number of safe deposit box categories with higher recommendation probabilities, and obtain the safe deposit box categories to be recommended.
[0131] In this embodiment, the top three safe deposit box categories with the highest recommendation probability can be selected as the safe deposit box categories to be recommended.
[0132] Step S210: Push the recommended safe deposit box category, receive the item storage evaluation feedback data, and update the safe deposit box category recommendation model based on the item storage evaluation feedback data.
[0133] Step S212: Obtain the business location corresponding to the safe deposit box category to be recommended and the location of the recommended safe deposit box category, and push a second preset number of business locations that are close to the location of the recommended safe deposit box category.
[0134] In practical applications, the process might involve obtaining authorization from the customer's geographic location after acquiring the recommended safe deposit box category. Simultaneously, it involves querying available safe deposit box locations corresponding to the category being recommended, calculating the distance between the customer's location and these locations, selecting the five closest locations, and then pushing these five locations to the client. This allows customers to complete safe deposit box transactions more quickly, saving time.
[0135] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.
[0136] Based on the same inventive concept, this application also provides a safe deposit box category recommendation device for implementing the safe deposit box category recommendation method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more safe deposit box category recommendation device embodiments provided below can be found in the limitations of the safe deposit box category recommendation method described above, and will not be repeated here.
[0137] In one embodiment, such as Figure 8 As shown, a safe deposit box category recommendation device is provided, including: a data acquisition module 810, a data query module 820, a recommendation probability determination module 830, and a safe deposit box category determination module 840, wherein:
[0138] The data acquisition module 810 is used to acquire the category and storage area of the items to be stored.
[0139] The data query module 820 is used to query climate characteristic data of the stored region.
[0140] The recommendation probability determination module 830 is used to determine the recommendation probability of the safe deposit box category based on the category of the item to be stored and the climate characteristic data of the storage area.
[0141] The safe deposit box category determination module 840 is used to determine the safe deposit box category to be recommended based on the recommendation probability of the safe deposit box category.
[0142] The aforementioned safe deposit box category recommendation device uses the category of the item to be stored and the storage location as benchmarks. It also considers the impact of climate differences in different regions on the storage requirements, further querying climate characteristic data for the storage location. Based on the category of the item to be stored and the climate characteristics of the storage location, it determines the recommendation probability of a safe deposit box category, and then determines the recommended safe deposit box category based on the recommendation probability. The entire solution not only considers the different requirements for box types due to differences in item categories, but also takes into account the potential damage caused by climate differences. By comprehensively considering the category of the item to be stored and the climate characteristics of the storage location, it provides targeted recommendations for safe deposit box categories that are more suitable for the item, achieving accurate recommendations. Furthermore, it can assist business personnel or guide customers in determining the final, more suitable safe deposit box category and effectively reduces the risk of damage to the stored items due to inaccurate safe deposit box category recommendations.
[0143] In one embodiment, the safe deposit box category determination module 840 is used to take the category of the item to be stored and the climate characteristics data of the storage area as input data, call the trained safe deposit box category recommendation model, and obtain the recommendation probability of the safe deposit box category;
[0144] The safe deposit box category recommendation model was trained based on climate characteristic data of the storage area and box type characteristic data.
[0145] like Figure 9 As shown, in one embodiment, the device further includes a model training module 822, used to acquire model sample data. The target model sample data includes climate distribution data and corresponding box type feature data of different regions. The target model sample data is preprocessed on the climate distribution data and corresponding box type feature data of different regions to obtain target model sample data. An initial safe deposit box category recommendation model is obtained. Based on the target model sample data, the initial safe deposit box category recommendation model is trained to obtain the safe deposit box category recommendation model.
[0146] In one embodiment, the model training module 822 is further used to perform numerical processing and clustering processing on the box-type feature data in sequence to obtain the classification results of the safe deposit box, to divide different regions according to the climate distribution data of different regions to obtain the climate feature data of different regions, and to collect the classification results of the safe deposit box and the corresponding climate feature data of different regions to obtain the target model sample data.
[0147] In one embodiment, the model training module 822 is further configured to train an initial safe deposit box category recommendation model based on the target model sample data, using a gradient descent algorithm and a dynamic learning rate, to obtain the safe deposit box category recommendation model.
[0148] In one embodiment, the safe deposit box category determination module 840 is further configured to sort the recommendation probabilities of safe deposit box categories, select a first preset number of safe deposit box categories with higher recommendation probabilities, and obtain the safe deposit box categories to be recommended.
[0149] like Figure 9 As shown, in one embodiment, the device further includes a model update module 824, which is used to push recommended safe deposit box categories, receive item storage evaluation feedback data, and update the safe deposit box category recommendation model based on the item storage evaluation feedback data.
[0150] like Figure 9 As shown, in one embodiment, the device further includes a business location push module 850, which is used to obtain the business location corresponding to the safe deposit box category to be recommended and the location of the recommended safe deposit box category object, and push a second preset number of business locations that are close to the location of the recommended safe deposit box category object.
[0151] Each module in the aforementioned recommended safe deposit box devices can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.
[0152] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores data such as box type characteristics and climate characteristics data for different regions. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a method for recommending safe deposit box categories.
[0153] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0154] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps in the above-described safe deposit box category recommendation method.
[0155] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method for recommending safe deposit box categories.
[0156] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described safe deposit box category recommendation method.
[0157] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0158] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0159] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.
Claims
1. A method for recommending safe deposit box categories, characterized in that, The method includes: Obtain the category and storage area of the items to be stored, wherein the category of the items to be stored includes documents, paintings and calligraphy, or precision instruments; Query the climate characteristic data of the stored region; The category of the items to be stored and the climate characteristics of the storage area are used as input data. A trained safe deposit box category recommendation model is called to obtain the recommendation probability of the safe deposit box category. The safe deposit box category recommendation model is trained based on the category of the items to be stored, the climate characteristics of the storage area, and the box type characteristics. The safe deposit box category is divided based on the box type characteristics, which include size, material, shockproof, moisture-proof, fireproof, and lock type. Based on the recommendation probability of the aforementioned safe deposit box categories, determine the safe deposit box categories to be recommended.
2. The method according to claim 1, characterized in that, Before using the category of the items to be stored and the climate characteristic data of the storage area as input data to call the trained safe deposit box category recommendation model to obtain the recommendation probability of the safe deposit box category, the method further includes: Obtain model sample data, which includes the category of items to be stored, climate distribution data of different regions, and corresponding box type feature data; Data preprocessing is performed on the categories of items to be stored, the climate distribution data of different regions, and the corresponding box type characteristic data to obtain target model sample data; Obtain the initial safe deposit box category recommendation model; Based on the target model sample data, the initial safe deposit box category recommendation model is trained to obtain the safe deposit box category recommendation model.
3. The method according to claim 2, characterized in that, The data preprocessing of the categories of items to be stored, the climate distribution data of different regions, and the box-type characteristic data to obtain target model sample data includes: The box type feature data are sequentially processed by numericalization and clustering to obtain the box category classification results; Based on the climate distribution data of different regions, the regions are divided to obtain climate characteristic data of different regions; The target model sample data is obtained by collecting the categories of the items to be stored, the classification results of the storage boxes, and the corresponding climate characteristic data of different regions.
4. The method according to claim 2, characterized in that, The process of training the initial safe deposit box category recommendation model based on the target model sample data to obtain the safe deposit box category recommendation model includes: Based on the target model sample data, the initial safe deposit box category recommendation model is trained using the gradient descent algorithm and dynamic learning rate to obtain the safe deposit box category recommendation model.
5. The method according to any one of claims 1 to 4, characterized in that, The step of determining the recommended safe deposit box category based on the recommendation probability of the safe deposit box category includes: The recommendation probabilities of the aforementioned safe deposit box categories are ranked. Select a first preset number of safe deposit box categories with a higher recommendation probability to obtain the safe deposit box categories to be recommended.
6. The method according to any one of claims 1 to 4, characterized in that, After determining the category of safe deposit box to be recommended, the method further includes: The recommended safe deposit box categories will be pushed out. Receive feedback data on item storage evaluation; The safe deposit box category recommendation model is updated based on the feedback data on the storage of the items.
7. The method according to claim 6, characterized in that, After pushing the recommended safe deposit box category, the method further includes: Obtain the business location corresponding to the safe deposit box category to be recommended, as well as the location of the recommended safe deposit box category; Push a second preset number of business locations that are close to the location of the recommended objects of the safe deposit box category.
8. A safe deposit box category recommendation device, characterized in that, The device includes: The data acquisition module is used to acquire the category and storage area of the items to be stored, wherein the category of the items to be stored includes documents, paintings and calligraphy, or precision instruments; The data query module is used to query the climate characteristic data of the stored area; The recommendation probability determination module is used to take the category of the item to be stored and the climate characteristic data of the storage area as input data, call the trained safe deposit box category recommendation model, and obtain the recommendation probability of the safe deposit box category; wherein, the safe deposit box category recommendation model is trained based on the category of the item to be stored, the climate characteristic data of the storage area, and the box type characteristic data, and the safe deposit box category is divided based on the box type characteristic data of the safe deposit box, including size, material, shockproof, moisture-proof, fireproof, and lock type; The safe deposit box category determination module is used to determine the safe deposit box category to be recommended based on the recommendation probability of the safe deposit box category.
9. The apparatus according to claim 8, characterized in that, The device further includes a model training module for acquiring model sample data, which includes the category of items to be stored, climate distribution data of different regions, and corresponding box type feature data. The module performs data preprocessing on the category of items to be stored, climate distribution data of different regions, and corresponding box type feature data to obtain target model sample data, acquires an initial safe deposit box category recommendation model, and trains the initial safe deposit box category recommendation model based on the target model sample data to obtain a safe deposit box category recommendation model.
10. The apparatus according to claim 9, characterized in that, The model training module is also used to perform numerical processing and clustering processing on the box-type feature data in sequence to obtain the classification results of the safe deposit box. Based on the climate distribution data of different regions, different regions are divided to obtain the climate feature data of different regions. The module collects the categories of the items to be stored, the classification results of the safe deposit box, and the corresponding climate feature data of different regions to obtain the target model sample data.
11. The apparatus according to claim 10, characterized in that, The model training module is also used to train the initial safe deposit box category recommendation model based on the target model sample data, using the gradient descent algorithm and dynamic learning rate, to obtain the safe deposit box category recommendation model.
12. The apparatus according to any one of claims 8 to 11, characterized in that, The safe deposit box category determination module is also used to sort the recommendation probabilities of the safe deposit box categories, select a first preset number of safe deposit box categories with higher recommendation probabilities, and obtain the safe deposit box categories to be recommended.
13. The apparatus according to any one of claims 8 to 11, characterized in that, The device also includes a model update module, which is used to push the recommended safe deposit box categories, receive item storage evaluation feedback data, and update the safe deposit box category recommendation model based on the item storage evaluation feedback data.
14. The apparatus according to claim 13, characterized in that, The device also includes a business location push module, which is used to obtain the business locations corresponding to the safe deposit box category to be recommended and the location of the recommended safe deposit box category object, and push a second preset number of business locations that are close to the location of the recommended safe deposit box category object.
15. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.
16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.
17. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.
Citation Information
Patent Citations
RFID tag-based multi-warehouse adaptive storage management method and apparatus
CN106651028A
Method for recommending packing box and apparatus therefor
KR102411587B1
