Floating information determination method and device, electronic equipment and storage medium
By determining the floating information intervals of historical users for historical products and matching the current user and product information to these intervals, the matching probability recommendation interval is calculated, and the cold start problem is solved and the accuracy of floating information recommendation is improved.
Patent Information
- Application Number
- CN202411898350.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-20
- Publication Date
- 2025-05-06
AI Technical Summary
The prior art faces cold start problems caused by the lack of historical data when predicting or recommending product floating information.
By determining multiple floating intervals corresponding to the floating information selected by historical users for historical products, and matching the current user's information and current product information with the historical data of these intervals, the matching probability of each interval is calculated, and the floating interval whose matching probability meets the requirements is recommended as the floating information recommended interval for the current product for the current user.
It effectively solves the cold start problem caused by the lack of historical data of a certain type of product or the lack of historical data of a certain type of product for a certain type of user, and improves the accuracy and reliability of floating information recommendations.
Smart Images
Figure CN119941289A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a method, device, electronic device and storage medium for determining floating information. Background Art
[0002] In the related art, the prediction or recommendation of product floating information (such as product profit margin, product production schedule and product delivery date, etc.) is determined based on the historical data of the product. For example, the profit margin prediction for product A is usually estimated based on the historical transaction data or historical order data of product A. However, for a certain type of product, the lack of historical data of the product or the lack of historical data of the product for a certain type of user will cause a cold start problem. Summary of the invention
[0003] In order to solve the above technical problems, the embodiments of the present application provide a method, device, electronic device and storage medium for determining floating information.
[0004] The technical solution of the embodiment of the present application is implemented as follows:
[0005] In a first aspect, an embodiment of the present application provides a method for determining floating information, including:
[0006] Determine multiple floating intervals corresponding to the floating information selected by the historical user for the historical product; the multiple floating intervals represent different floating intervals to which multiple pieces of historical data belong, each piece of the multiple pieces of historical data is used to represent the selection of the floating information of the historical product by the historical user, and each piece of the historical data includes information of the historical user and information of the historical product;
[0007] Matching the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval respectively, to obtain the matching probability corresponding to each floating interval respectively;
[0008] A floating interval in which the matching probability meets the requirement is determined as a first floating interval, so as to use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0009] In some embodiments, the matching of the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval to obtain the matching probabilities corresponding to each floating interval includes:
[0010] Obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval;
[0011] Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by utilizing the ridge regression models corresponding to the floating intervals.
[0012] In some embodiments, after determining that the floating interval that meets the requirements of the matching probability is the first floating interval and using the first floating interval as the floating information recommendation interval of the current product for the current user, the method further includes:
[0013] Obtaining feedback information of the current user with respect to the first floating interval;
[0014] When it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, the first floating interval is updated based on the feedback information to obtain an updated first floating interval, so as to use the updated first floating interval as a floating information recommendation interval of the current product for the current user.
[0015] In some embodiments, updating the first floating interval based on the feedback information to obtain the updated first floating interval includes:
[0016] Based on the feedback information, updating the parameters of the ridge regression model corresponding to the first floating interval;
[0017] Based on the information of the current user and the information of the current product, re-determine the matching probability corresponding to the first floating interval by using the ridge regression model corresponding to the first floating interval after parameter update;
[0018] From the matching probability corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals among the multiple floating intervals except the first floating interval, a floating interval whose matching probability meets the requirements is determined as the updated first floating interval.
[0019] In some embodiments, determining a plurality of floating intervals corresponding to the floating information selected by historical users for historical products includes:
[0020] Determine the historical floating information corresponding to each of the plurality of historical data, wherein the historical floating information is the floating information selected by the historical user for the historical product;
[0021] The floating information range is divided into the plurality of floating intervals based on the probability density and the specific probability interval of each of the historical floating information within the floating information range.
[0022] In some embodiments, before determining a plurality of floating intervals corresponding to the floating information selected by the historical user for the historical product, the method further includes:
[0023] Performing logical reasoning on the selection of the floating information of the historical product by the historical user represented by each of the plurality of historical data, to obtain derivative data corresponding to each of the historical data;
[0024] The derivative data corresponding to each piece of the historical data are respectively incorporated into the plurality of pieces of historical data.
[0025] In some embodiments, the user information includes at least one of the following: identity information of the user, industry information of the user, geographic location information of the user, product transportation method information that the user is willing to accept, and user willingness to purchase products;
[0026] The product information includes at least one of the following: brand information of the product, order information of the product, cost information of the product, service information bound to the product, and market share information of the product.
[0027] In a second aspect, an embodiment of the present application provides a floating information determination device, including:
[0028] A first determination module is used to determine a plurality of floating intervals corresponding to the floating information selected by the historical user for the historical product; the plurality of floating intervals represent different floating intervals to which a plurality of historical data belong, each of the plurality of historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes information of the historical user and information of the historical product;
[0029] A matching module, used to match the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval, respectively, to obtain the matching probability corresponding to each floating interval;
[0030] The second determination module is used to determine a floating interval in which the matching probability meets the requirements as a first floating interval, so as to use the first floating interval as a floating information recommendation interval for the current product to the current user.
[0031] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the memory is used to store executable data instructions; when the processor is used to execute the executable data instructions stored in the memory, the method for determining floating information as described in the first aspect is implemented.
[0032] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the method for determining floating information as described in the first aspect is implemented. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present application, a brief introduction will be given below to the drawings required for use in the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0034] Figure 1 One of the flow charts of a method for determining floating information provided in an embodiment of the present application;
[0035] Figure 2 A second flow chart of a method for determining floating information provided in an embodiment of the present application;
[0036] Figure 3 A third flow chart of a method for determining floating information provided in an embodiment of the present application;
[0037] Figure 4 A fourth flowchart of a method for determining floating information provided in an embodiment of the present application;
[0038] Figure 5 A fifth flow chart of a method for determining floating information provided in an embodiment of the present application;
[0039] Figure 6 A schematic diagram of a floating interval division result provided in an embodiment of the present application;
[0040] Figure 7 A schematic diagram of a principle for determining floating information provided in an embodiment of the present application;
[0041] Figure 8 A schematic diagram of the structure of a floating information determination device provided in an embodiment of the present application;
[0042] Fig. 9 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0043] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the embodiments of the present application.
[0044] It should be noted that in the description of the embodiments of the present application, the terms "first", "second", etc. are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of the same type, and the number of objects is not limited. For example, the first object can be one or more.
[0045] In order to facilitate a clearer understanding of the embodiments of the present application, some relevant technical knowledge is first introduced as follows.
[0046] Take the product profit margin estimation scenario as an example. In this scenario, the relevant technology usually estimates the profit margin based on the historical transaction data of a certain type of product after receiving a customer's request order for the product, and obtains a profit margin estimate (including a conservative profit margin estimate, a moderate profit margin estimate, and an aggressive profit margin estimate), so that the salesperson can use the profit margin estimate as a reference for at least one round of communication with the customer, so as to reach a transaction between the product and the customer with the least possible reduction in profit margin. However, when the product to be traded is a new product that has just entered the market, the profit margin of the product cannot be estimated due to the lack of historical data for the product, resulting in a cold start problem.
[0047] In order to overcome the above-mentioned defects of the related art, the embodiments of the present application propose a method, device, electronic device and storage medium for determining floating information, which can match the information of the current user and the current product with the information of the historical user and the information of the historical product in the historical data of each floating interval based on the different floating intervals to which the existing historical data for historical products and historical users belong, and use the floating interval with the matching probability meeting the requirements as the floating information recommendation interval of the current product for the current user, effectively improving the cold start problem caused by the lack of historical data of a certain type of product or the lack of historical data of a certain type of product for a certain type of user.
[0048] It should be noted that, in the scenario of product profit margin estimation, the floating information in the embodiment of the present application is the product profit margin information.
[0049] In conjunction with the drawings in the embodiments of the present application, a method, device, electronic device and storage medium for determining floating information provided in the embodiments of the present application are exemplarily introduced below.
[0050] Figure 1 One of the flow charts of a method for determining floating information provided in an embodiment of the present application is as follows: Figure 1 As shown, the method includes:
[0051] S101. Determine multiple floating intervals corresponding to the floating information selected by historical users for historical products; the multiple floating intervals represent different floating intervals to which multiple historical data belong, each of the multiple historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product.
[0052] It should be noted that floating information is a dynamically changing data, which may include but is not limited to product profit margin information (or price information), product production date information, product delivery date information, etc. Taking the product profit margin estimation scenario as an example, the floating information selected by historical users for historical products is the profit margin information.
[0053] In some embodiments, multiple floating intervals corresponding to the floating information selected by the historical user for the historical product can be determined based on multiple historical data. Taking the product profit margin estimation scenario as an example, the profit margin information selected by the historical user for the historical product can be determined based on multiple historical product order data or multiple historical product transaction data.
[0054] It should be noted that each of the multiple pieces of historical data can represent the selection of the floating information of the historical product by the historical user. For example, if the historical user a purchased the historical product c at the profit rate b, the profit rate information selected by the historical user a for the historical product c is the profit rate b.
[0055] It should be noted that each piece of historical data includes historical user information and historical product information. Among them, historical user information refers to user data of historical users who have interacted or traded with a product in the past period of time. This information can include various types, which helps to know or understand historical user behavior, historical user preferences and historical user needs. Historical product information refers to a series of data and records about products produced, sold or existed in the past. This information helps to know or understand the evolution of products, market performance and user feedback.
[0056] In some embodiments, each piece of historical data may be converted into a feature vector, wherein the feature vector corresponding to each piece of historical data includes a feature vector corresponding to the historical user information and a feature vector corresponding to the historical product information, and may also include a feature vector that can characterize the selection of the historical user for the floating information of the historical product.
[0057] S102: Match the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval, respectively, to obtain the matching probabilities corresponding to each floating interval.
[0058] In the embodiment of the present application, when the order information of the current user for the current product is received, the matching probability of the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval can be calculated. The matching probability can be the average of the matching probability of the current user information and the historical user information and the matching probability of the current product information and the historical product information.
[0059] In some embodiments, the current user information and the current product information can be converted into feature vectors respectively, and then the feature vector corresponding to the current user information and the feature vector corresponding to the historical user information in the historical data of each floating interval are matched. At the same time, the feature vector corresponding to the current product information and the feature vector corresponding to the historical product information in the historical data of each floating interval are matched. Finally, the average of the two matching probabilities corresponding to each floating interval is calculated to obtain the final matching probability corresponding to each floating interval.
[0060] In some embodiments, the user information includes at least one of the following: identity information of the user, industry information of the user, geographic location information of the user, product transportation method information that the user is willing to accept, and user willingness to purchase products;
[0061] The product information includes at least one of the following: brand information of the product, order information of the product, cost information of the product, service information bound to the product, and market share information of the product.
[0062] It should be noted that the information of the user in the embodiment of the present application may include the information of the historical user and the information of the current user. That is, the information of the historical user in the historical data may include at least one of the following: the identity information of the historical user, the industry information of the historical user, the geographical location information of the historical user, the product transportation method information that the historical user is willing to accept, and the willingness information of the historical user to purchase the historical product; the information of the current user may include at least one of the following: the identity information of the current user, the industry information of the current user, the geographical location information of the current user, the product transportation method information that the current user is willing to accept, and the willingness information of the current user to purchase the current product. Similarly, the information of the product in the embodiment of the present application may include the information of the historical product and the information of the current product. That is, the information of the historical product in the historical data may include at least one of the following: the brand information of the historical product, the order information of the historical product, the cost information of the historical product, the service information bound to the historical product, and the market share information of the historical product; the information of the current product may include at least one of the following: the brand information of the current product, the order information of the current product, the cost information of the current product, the service information bound to the current product, and the market share information of the current product.
[0063] S103: Determine a floating interval in which the matching probability meets the requirement as a first floating interval, and use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0064] It should be noted that the floating interval in which the matching probability meets the requirements can be the floating interval corresponding to the maximum matching probability or the floating interval corresponding to the matching probability being greater than a specific value. As to which floating intervals correspond to matching probabilities that meet the requirements, specific requirements can be adaptively set based on actual applications or actual needs, and the embodiments of the present application do not specifically limit this.
[0065] In the embodiment of the present application, after obtaining the matching probabilities corresponding to the respective floating intervals, the floating intervals with matching probabilities that meet the requirements can be used as the floating information recommendation intervals of the current product for the current user. The floating information recommendation intervals can be used to display to the current user, or to recommend to the salesperson of the current product, so that the salesperson can use the floating information recommendation intervals as a reference to communicate with the current user about the transaction of the current product.
[0066] It can be understood that the method for determining floating information provided in the embodiment of the present application first determines multiple floating intervals corresponding to the floating information selected by historical users for historical products, wherein the multiple floating intervals represent different floating intervals to which multiple historical data belong, and each of the multiple historical data is used to represent the selection of floating information of historical products by historical users, and each of the multiple historical data includes information of historical users and information of historical products, and then the information of the current user and the information of the current product are matched with the information of historical users and the information of historical products in the historical data of each floating interval, respectively, to obtain the matching probabilities corresponding to each floating interval, and then determine the floating interval whose matching probability meets the requirements. The first floating interval is used as the floating information recommendation interval of the current product for the current user. In this way, even if there is a lack of historical data for the current product and / or the current user, the embodiment of the present application can still match the information of the current user and the information of the current product with the information of the historical users and the information of the historical products in the historical data of each floating interval based on the different floating intervals to which the existing historical data for the historical products and historical users belong, and use the floating interval with the matching probability meeting the requirements as the floating information recommendation interval of the current product for the current user, effectively improving the cold start problem caused by the lack of historical data of a certain type of product or the lack of historical data of a certain type of product for a certain type of user.
[0067] In some embodiments, the matching of the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval to obtain the matching probabilities corresponding to each floating interval includes:
[0068] Obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval;
[0069] Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by utilizing the ridge regression models corresponding to the floating intervals.
[0070] It should be noted that Ridge Regression is an extension of linear regression. By adding the L2 regularization term to the loss function, it can effectively deal with the multicollinearity problem between independent variables, so that the model can still maintain stability and accuracy when facing highly correlated features.
[0071] It should be noted that since the historical data belonging to each floating interval have highly correlated characteristics, the embodiment of the present application establishes a ridge regression model corresponding to each floating interval based on the historical data belonging to each floating interval, and then inputs the information of the current user and the information of the current product into the ridge regression model corresponding to each floating interval, so as to obtain the matching probability corresponding to each floating interval more accurately.
[0072] In some embodiments, for any second floating interval in each floating interval, the feature vectors corresponding to the m pieces of historical data belonging to the second floating interval may be formed into a matrix D a , the matrix D a The dimension is m×d, where d is the dimension of the feature vector corresponding to each piece of historical data; the selection conditions (0 or 1) of the historical users for the floating information of the historical products represented by the m pieces of historical data are combined into a vector C a The expression of the ridge regression model is C a =X a T ·θ a Among them, X a is the independent variable, θ a is the parameter of the ridge regression model, and its dimension is d. According to the expression of the ridge regression model, we can get θ a =(D a T D a +I) -1 D a T C a Let A a =D a T D a +I,b a =D a T C a , then θ a =A a -1 b a . Where I is the identity matrix.
[0073] It can be understood that the matrix D composed of the eigenvectors corresponding to the historical data belonging to each floating interval a Different, so the parameters θ of the ridge regression model corresponding to each floating interval a different.
[0074] For example, Figure 2 A second flow chart of a method for determining floating information provided in an embodiment of the present application is as follows: Figure 2 As shown, the method includes:
[0075] S201. Determine multiple floating intervals corresponding to the floating information selected by historical users for historical products; the multiple floating intervals represent different floating intervals to which multiple historical data belong, each of the multiple historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product.
[0076] S202, obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval.
[0077] S203 : Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by using the ridge regression models corresponding to the floating intervals.
[0078] S204: Determine a floating interval in which the matching probability meets the requirement as a first floating interval, and use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0079] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.
[0080] It can be understood that the embodiment of the present application can more accurately obtain the matching probability corresponding to each floating interval by using the ridge regression model corresponding to each floating interval established based on the historical data belonging to each floating interval based on the information of the current user and the current product.
[0081] In some embodiments, after determining that the floating interval that meets the requirements of the matching probability is the first floating interval and using the first floating interval as the floating information recommendation interval of the current product for the current user, the method further includes:
[0082] Obtaining feedback information of the current user with respect to the first floating interval;
[0083] When it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, the first floating interval is updated based on the feedback information to obtain an updated first floating interval, so as to use the updated first floating interval as a floating information recommendation interval of the current product for the current user.
[0084] It should be noted that the feedback information of the current user on the first floating interval can represent the current user's satisfaction with the first floating interval. For example, if the feedback information of the current user on the first floating interval is that the floating value of the first floating interval is too high, it can represent that the current user is dissatisfied with the first floating interval.
[0085] In an embodiment of the present application, after the first floating interval is recommended to the current user as the floating information recommendation interval of the current product for the current user, feedback information of the current user regarding the first floating interval can be obtained, and then based on the feedback information, if it is determined that the current user is not satisfied with the first floating interval, the first floating interval can be updated (i.e., the first floating interval is re-determined from various floating intervals), and then the updated first floating interval is used as the floating information recommendation interval of the current product for the current user, that is, the updated first floating interval can be used to display to the current user, or the updated first floating interval can be used to provide to sales personnel, so that the sales personnel will use the updated first floating interval as a reference to communicate with the current user again regarding the transaction of the current product.
[0086] For example, Figure 3 A third flow chart of a method for determining floating information provided in an embodiment of the present application is as follows: Figure 3 As shown, the method includes:
[0087] S301. Determine multiple floating intervals corresponding to the floating information selected by historical users for historical products; the multiple floating intervals represent different floating intervals to which multiple historical data belong, each of the multiple historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product.
[0088] S302, obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval in each of the floating intervals is established based on the historical data belonging to the second floating interval.
[0089] S303 : Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by using the ridge regression models corresponding to the floating intervals.
[0090] S304: Determine a floating interval in which the matching probability meets the requirement as a first floating interval, and use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0091] S305: Obtain feedback information of the current user regarding the first floating interval.
[0092] S306. When it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, based on the feedback information, the first floating interval is updated to obtain an updated first floating interval, so as to use the updated first floating interval as a floating information recommendation interval of the current product for the current user.
[0093] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.
[0094] It can be understood that the embodiment of the present application can obtain the user's sensitivity and acceptance of the floating information of the current product based on the feedback information of the current user on the first floating interval, and then update the first floating interval based on the feedback information, which is equivalent to optimizing the floating interval corresponding to the floating information of the current product, obtaining a more reasonable floating interval to recommend to the current user, thereby enhancing the attractiveness of the current product to the current user.
[0095] In some embodiments, updating the first floating interval based on the feedback information to obtain the updated first floating interval includes:
[0096] Based on the feedback information, updating the parameters of the ridge regression model corresponding to the first floating interval;
[0097] Based on the information of the current user and the information of the current product, re-determine the matching probability corresponding to the first floating interval by using the ridge regression model corresponding to the first floating interval after parameter update;
[0098] From the matching probability corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals among the multiple floating intervals except the first floating interval, a floating interval whose matching probability meets the requirements is determined as the updated first floating interval.
[0099] In an embodiment of the present application, when it is determined that the current user is dissatisfied with the first floating interval, the parameters of the ridge regression model corresponding to the first floating interval can be updated based on the feedback information of the current user regarding the first floating interval, so that the matching probability calculated by the ridge regression model corresponding to the first floating interval based on the information of the current user and the information of the current product is reduced, and then the floating interval whose matching probability meets the requirements is determined as the updated first floating interval from the matching probability corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals in the multiple floating intervals except the first floating interval.
[0100] In some embodiments, the matching probability may be calculated based on the following formula (1):
[0101] P a =X a T ·θ a +a·(X a T A a -1 X a ) 1 / 2 (1)
[0102] Among them, P a represents the matching probability, X a T ·θ a represents the matching probability calculated based on the ridge regression model, α is a specific parameter (which can be set adaptively based on actual applications), a·(X a T A a -1 X a ) 1 / 2 Represents the uncertainty probability value for the matching probability calculated based on the ridge regression model. It should be noted that, since the amount of historical data may be insufficient, the matching probability calculated by the ridge regression model established based on the insufficient amount of historical data has uncertainty factors. Therefore, the embodiment of the present application obtains the final matching probability by summing the matching probability calculated based on the ridge regression model and its corresponding uncertainty probability value, which can improve the accuracy of the matching probability calculation.
[0103] It can be understood that if the ridge regression model is used to calculate the matching probability based on the current user information and the current product information, the current user information and the current product information can be converted into a feature vector, and then the feature vector is substituted into X in the above formula (1). a , the matching probability P can be calculated by the above formula (1) a .
[0104] In some embodiments, if the feedback information r of the current user for the first floating interval is obtained, and the feedback information r indicates that the current user is dissatisfied with the first floating interval, then the parameter θ of the ridge regression model corresponding to the first floating interval can be calculated based on the feedback information r in combination with the following formulas (2), (3) and (4): a Update and obtain the updated parameter θ a ':
[0105] A a '=A a +X a ·X a T(2)
[0106] b a '=b a +r·X a (3)
[0107] θ a '=A a ' -1 b a '(4)
[0108] Among them, X a As the independent variable, the feature vectors corresponding to the previous user information and the current product information can be substituted into X in equations (2) and (3): a , to calculate the updated A a ' and b a '.
[0109] For example, Figure 4 A fourth flow chart of a method for determining floating information provided in an embodiment of the present application is as follows: Figure 4 As shown, the method includes:
[0110] S401. Determine multiple floating intervals corresponding to the floating information selected by the historical user for the historical product; the multiple floating intervals represent different floating intervals to which multiple historical data belong, each of the multiple historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product.
[0111] S402, obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval.
[0112] S403 : Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by using the ridge regression models corresponding to the floating intervals.
[0113] S404: Determine a floating interval in which the matching probability meets the requirement as a first floating interval, and use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0114] S405: Obtain feedback information of the current user regarding the first floating interval.
[0115] S406: When it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, based on the feedback information, update the parameters of the ridge regression model corresponding to the first floating interval.
[0116] S407: Based on the information of the current user and the information of the current product, the matching probability corresponding to the first floating interval is re-determined using the ridge regression model corresponding to the first floating interval after parameter update.
[0117] S408. Determine a floating interval whose matching probability meets the requirements from the matching probabilities corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals among the multiple floating intervals except the first floating interval as the updated first floating interval, so as to use the updated first floating interval as the floating information recommendation interval of the current product for the current user.
[0118] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.
[0119] It should be noted that the multi-armed bandit in the contextual multi-armed bandit (Contextual Multi-Armed Bandit, Contextual MAB) algorithm in online learning technology refers to a multi-armed bandit with multiple optional actions (or called "arms", "bandit levers", etc.), and each action will give a reward (or called "revenue", "return", etc.) when it is selected. "Context" refers to additional information that can be observed before selecting an action, which can be used to assist decision-making to increase the sum of long-term rewards. The embodiment of the present application modifies the Contextual MAB algorithm to adapt to the floating information determination business of the product. Taking the product profit margin estimation scenario as an example, the embodiment of the present application can first classify historical data and match user information and product information to obtain a first profit margin range for recommendation to the current user, and then obtain the current user's feedback information on the first profit margin range (equivalent to the "context" in the multi-armed bandit algorithm with context, that is, additional information that can be observed before selecting an action), and then use the feedback information to update the parameters of the ridge regression model corresponding to the first profit margin range online, and finally, based on the ridge regression model with updated parameters, update the first profit margin range to recommend it to the current user again, thereby increasing the attractiveness of the current product to the current user and thus increasing the success rate of the transaction.
[0120] In some embodiments, determining a plurality of floating intervals corresponding to the floating information selected by historical users for historical products includes:
[0121] Determine the historical floating information corresponding to each of the plurality of historical data, wherein the historical floating information is the floating information selected by the historical user for the historical product;
[0122] The floating information range is divided into the plurality of floating intervals based on the probability density and the specific probability interval of each of the historical floating information within the floating information range.
[0123] In an embodiment of the present application, taking the product profit margin prediction scenario as an example, historical profit margin information corresponding to multiple historical data can be obtained, and then based on the probability density of the historical profit margin information corresponding to the multiple historical data within the profit margin range (0 to 1) and a specific probability interval (for example, 10%), the profit margin range (0 to 1) is divided into multiple (for example, 10 (determined based on the specific probability interval)) profit margin intervals, so that the multiple historical data are classified into different profit margin intervals.
[0124] In some embodiments, where the probability density is large, the intervals of the floating intervals can be divided into smaller intervals; and where the probability density is small, the intervals of the floating intervals can be divided into larger intervals. How to divide the floating intervals according to the probability density can be adaptively set based on the actual application, and the embodiments of the present application do not specifically limit this.
[0125] It should be noted that the specific probability interval can be adaptively set based on actual applications, and the embodiments of the present application do not specifically limit this.
[0126] For example, Figure 5 A fifth flow chart of a method for determining floating information provided in an embodiment of the present application is as follows: Figure 5 As shown, the method includes:
[0127] S501: Determine historical floating information corresponding to each of a plurality of historical data, where the historical floating information is floating information selected by historical users for historical products.
[0128] S502: Divide the floating information range into a plurality of floating intervals based on the probability density and specific probability interval of each of the historical floating information within the floating information range.
[0129] S503: Match the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval, respectively, to obtain the matching probabilities corresponding to each floating interval.
[0130] S504: Determine a floating interval in which the matching probability meets the requirement as a first floating interval, and use the first floating interval as a floating information recommendation interval of the current product for the current user.
[0131] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.
[0132] For example, Figure 6 A schematic diagram of a floating interval division result provided in an embodiment of the present application is shown in FIG. Figure 6 As shown in the figure, the dotted intervals represent the profit margin intervals. It can be seen that the profit margin of the product conforms to the normal distribution, that is, most of the profit margins are concentrated between 0.2-0.5, and very few of the profit margins are concentrated around 0 and 1. According to the probability interval of 10%, 10 profit margin intervals can be divided, and according to the probability density of profit margin concentration, the intervals divided where the profit margin is concentrated are smaller, and the intervals divided where the profit margin is not concentrated are larger.
[0133] It can be understood that in order to assign multiple historical data to different floating intervals, the embodiment of the present application divides the floating information range into multiple floating intervals based on the probability density of each historical floating information within the floating information range, so that the historical data can be classified more accurately, which helps to improve the accuracy of the floating information recommendation interval.
[0134] In some embodiments, before determining a plurality of floating intervals corresponding to the floating information selected by the historical user for the historical product, the method further includes:
[0135] Performing logical reasoning on the selection of the floating information of the historical product by the historical user represented by each of the plurality of historical data, to obtain derivative data corresponding to each of the historical data;
[0136] The derivative data corresponding to each piece of the historical data are respectively incorporated into the plurality of pieces of historical data.
[0137] For example, for a certain piece of historical data, if historical user a purchased historical product c at profit rate b, then according to the purchase logic, historical user a will also purchase this product between profit rates 0 and b (this data is the derived data); for another piece of historical data, if historical user a did not purchase historical product c at profit rate b, then according to the purchase logic, historical user a will not purchase historical product c between profit rates b and 1 (this data is the derived data). Based on this logical reasoning, the historical data can be expanded or enhanced to make the historical data more diverse and rich, thereby improving the accuracy of the floating information recommendation interval.
[0138] For example, Figure 7 A schematic diagram of a principle for determining floating information provided in an embodiment of the present application, such as Figure 7 As shown in FIG. 1 , arm1 to arm10 represent the multiple floating intervals divided. After inputting the feature vector corresponding to the current user information and the feature vector corresponding to the current product information into the ridge regression model corresponding to each arm (floating interval), the matching probability corresponding to each floating interval can be obtained. Figure 7 As shown, after arm6 whose matching probability meets the requirements obtained by the initial calculation is recommended to the current user, feedback information of the current user for arm6 is obtained, and the parameters of the ridge regression model corresponding to arm6 are updated based on the feedback information. After the update, it is re-determined that the matching probability corresponding to arm5 meets the requirements.
[0139] It is understandable that the embodiments of the present application can interactively optimize and adjust the floating information recommendation interval based on the user's feedback information to obtain a more reasonable floating information recommendation interval for the current user and the current product; moreover, the ridge regression model corresponding to each floating interval can be learned online, and finally a customized ridge regression model for a certain type of user and / or a certain type of product can be obtained. In addition, in the process of solving the ridge regression model, an optimization algorithm is usually used to minimize the loss function. These optimization algorithms include but are not limited to the gradient descent method and the least squares method. These algorithms can ensure that the ridge regression model converges to a stable solution within a guaranteed time complexity to obtain a floating interval that satisfies the user, and avoid the situation where the user loses patience due to multiple recommendations of floating intervals, which leads to transaction failures.
[0140] The following is a description of a floating information determination device provided in an embodiment of the present application. The floating information determination device described below and the floating information determination method described above can correspond to each other.
[0141] Figure 8 A schematic diagram of a floating information determination device provided in an embodiment of the present application is shown in FIG. Figure 8 As shown, the device includes: a first determination module 810, a matching module 820, and a second determination module 830; wherein:
[0142] The first determination module 810 is used to determine a plurality of floating intervals corresponding to the floating information selected by the historical user for the historical product; the plurality of floating intervals represent different floating intervals to which a plurality of historical data belong, each of the plurality of historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product;
[0143] A matching module 820, for matching the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval, to obtain matching probabilities corresponding to each floating interval;
[0144] The second determination module 830 is configured to determine a floating interval in which the matching probability meets the requirement as a first floating interval, so as to use the first floating interval as a floating information recommendation interval for the current product to the current user.
[0145] The floating information determination device provided by the embodiment of the present application first determines a plurality of floating intervals corresponding to the floating information selected by the historical users for the historical products, wherein the plurality of floating intervals represent different floating intervals to which the plurality of historical data belong, and each of the plurality of historical data is used to represent the selection of the floating information of the historical products by the historical users, and each of the historical data includes the information of the historical users and the information of the historical products, and then respectively matches the information of the current user and the information of the current product with the information of the historical users and the information of the historical products in the historical data of each floating interval, obtains the matching probabilities corresponding to the respective floating intervals, and then determines the floating interval whose matching probability meets the requirements as the first floating interval. The floating interval is used to take the first floating interval as the floating information recommendation interval of the current product for the current user. In this way, even if there is a lack of historical data for the current product and / or the current user, the embodiment of the present application can still match the information of the current user and the information of the current product with the information of the historical user and the information of the historical product in the historical data of each floating interval based on the different floating intervals to which the existing historical data for the historical products and historical users belong, and take the floating interval with the matching probability meeting the requirements as the floating information recommendation interval of the current product for the current user, which effectively improves the cold start problem caused by the lack of historical data of a certain type of product or the lack of historical data of a certain type of product for a certain type of user.
[0146] In some embodiments, the matching module 820 includes:
[0147] An obtaining unit, used for obtaining the ridge regression models corresponding to the respective floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval;
[0148] The first determining unit is used to determine the matching probability corresponding to each of the floating intervals based on the information of the current user and the information of the current product and by using the ridge regression model corresponding to each of the floating intervals.
[0149] In some embodiments, the apparatus further comprises:
[0150] an obtaining module, configured to obtain feedback information of the current user with respect to the first floating interval;
[0151] An updating module is used to update the first floating interval based on the feedback information when it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, so as to obtain an updated first floating interval and use the updated first floating interval as a floating information recommendation interval for the current product for the current user.
[0152] In some embodiments, the update module includes:
[0153] an updating unit, configured to update the parameters of the ridge regression model corresponding to the first floating interval based on the feedback information;
[0154] A second determining unit is configured to re-determine the matching probability corresponding to the first floating interval based on the information of the current user and the information of the current product and by using the ridge regression model corresponding to the first floating interval after parameter update;
[0155] The third determination unit is used to determine a floating interval whose matching probability meets the requirements as the updated first floating interval from the matching probability corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals in the multiple floating intervals except the first floating interval.
[0156] In some embodiments, the first determining module 810 includes:
[0157] A fourth determining unit, configured to determine historical floating information corresponding to each of the plurality of historical data, wherein the historical floating information is floating information selected by the historical user for the historical product;
[0158] The division unit is used to divide the floating information range into the multiple floating intervals based on the probability density and specific probability interval of each historical floating information within the floating information range.
[0159] In some embodiments, the apparatus further comprises:
[0160] An inference module, configured to perform logical inference on the selection of the floating information of the historical product by the historical user represented by each of the plurality of historical data, and obtain derivative data corresponding to each of the historical data;
[0161] The merging module is used to merge the derivative data corresponding to each of the historical data into the multiple historical data.
[0162] In some embodiments, the user information includes at least one of the following: identity information of the user, industry information of the user, geographic location information of the user, product transportation method information that the user is willing to accept, and user willingness to purchase products;
[0163] The product information includes at least one of the following: brand information of the product, order information of the product, cost information of the product, service information bound to the product, and market share information of the product.
[0164] It should be noted here that the above-mentioned floating information determination device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned floating information determination method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as those of the method embodiment will not be described in detail here.
[0165] Fig. 9 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present application, such as Fig. 9 As shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930 and a communication bus 940, wherein the processor 910, the communication interface 920 and the memory 930 communicate with each other via the communication bus 940. The processor 910 may call executable data instructions stored in the memory 930 to execute part or all of the steps in the method for determining floating information provided in the above embodiments.
[0166] In addition, the executable data instructions stored in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the embodiment of the present application can be essentially or partly embodied in the form of a software product that contributes to the relevant technology. The software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.
[0167] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, part or all of the steps in the method for determining floating information provided in the above embodiments are implemented.
[0168] An embodiment of the present application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute some or all of the steps in the floating information determination method provided in the above embodiments.
[0169] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.
[0170] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the embodiments of the present application may adopt the form of hardware embodiments, software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.
[0171] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0172] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0173] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0174] The above description is merely an optional embodiment of the present application and is not intended to limit the protection scope of the present application.
Claims
1. A method for determining floating information, comprising: Determine multiple floating intervals corresponding to the floating information selected by the historical user for the historical product; the multiple floating intervals represent different floating intervals to which multiple pieces of historical data belong, each piece of the multiple pieces of historical data is used to represent the selection of the floating information of the historical product by the historical user, and each piece of the historical data includes information of the historical user and information of the historical product; Matching the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval respectively, to obtain the matching probability corresponding to each floating interval respectively; A floating interval in which the matching probability meets the requirement is determined as a first floating interval, so as to use the first floating interval as a floating information recommendation interval of the current product for the current user.
2. The method for determining floating information according to claim 1, wherein the current user information and the current product information are matched with the historical user information and the historical product information in the historical data of each floating interval to obtain the matching probability corresponding to each floating interval, including: Obtaining ridge regression models corresponding to each of the floating intervals; the ridge regression model corresponding to any second floating interval among the floating intervals is established based on the historical data belonging to the second floating interval; Based on the information of the current user and the information of the current product, the matching probabilities corresponding to the floating intervals are determined by utilizing the ridge regression models corresponding to the floating intervals.
3. According to the method for determining floating information of claim 2, the floating interval in which the matching probability meets the requirement is determined to be the first floating interval, and after the first floating interval is used as the floating information recommendation interval of the current product for the current user, the method further comprises: Obtaining feedback information of the current user with respect to the first floating interval; When it is determined that the feedback information indicates that the current user is dissatisfied with the first floating interval, the first floating interval is updated based on the feedback information to obtain an updated first floating interval, so as to use the updated first floating interval as a floating information recommendation interval of the current product for the current user.
4. The method for determining floating information according to claim 3, wherein the updating of the first floating interval based on the feedback information to obtain the updated first floating interval comprises: Based on the feedback information, updating the parameters of the ridge regression model corresponding to the first floating interval; Based on the information of the current user and the information of the current product, re-determine the matching probability corresponding to the first floating interval by using the ridge regression model corresponding to the first floating interval after parameter update; From the matching probability corresponding to the re-determined first floating interval and the matching probabilities corresponding to the remaining floating intervals among the multiple floating intervals except the first floating interval, a floating interval whose matching probability meets the requirements is determined as the updated first floating interval.
5. The method for determining floating information according to claim 1, wherein the step of determining a plurality of floating intervals corresponding to the floating information selected by historical users for historical products comprises: Determine the historical floating information corresponding to each of the plurality of historical data, wherein the historical floating information is the floating information selected by the historical user for the historical product; The floating information range is divided into the plurality of floating intervals based on the probability density and the specific probability interval of each of the historical floating information within the floating information range.
6. The method for determining floating information according to claim 1, before determining a plurality of floating intervals corresponding to the floating information selected by historical users for historical products, the method further comprises: Performing logical reasoning on the selection of the floating information of the historical product by the historical user represented by each of the plurality of historical data, to obtain derivative data corresponding to each of the historical data; The derivative data corresponding to each piece of the historical data are incorporated into the plurality of pieces of historical data.
7. According to the method for determining floating information according to any one of claims 1 to 6, the user information includes at least one of the following: the user's identity information, the user's industry information, the user's geographical location information, the user's product transportation method information that the user is willing to accept, and the user's willingness to purchase products; The product information includes at least one of the following: brand information of the product, order information of the product, cost information of the product, service information bound to the product, and market share information of the product.
8. A floating information determination device, comprising: A first determination module is used to determine a plurality of floating intervals corresponding to the floating information selected by the historical user for the historical product; The multiple floating intervals represent different floating intervals to which the multiple historical data belong, each of the multiple historical data is used to represent the selection of the floating information of the historical product by the historical user, and each of the historical data includes the information of the historical user and the information of the historical product; A matching module, used to match the current user information and the current product information with the historical user information and the historical product information in the historical data of each floating interval, respectively, to obtain the matching probability corresponding to each floating interval; The second determination module is used to determine a floating interval in which the matching probability meets the requirements as a first floating interval, so as to use the first floating interval as a floating information recommendation interval for the current product to the current user.
9. An electronic device, comprising: A memory for storing executable data instructions; A processor, configured to implement the method for determining floating information according to any one of claims 1 to 7 when executing executable data instructions stored in the memory.
10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method for determining floating information according to any one of claims 1 to 7 is implemented.