Online fruit mall integral exchange method and system based on freshness dynamic pricing

Abstract

This invention discloses a method and system for redeeming points in an online fruit marketplace based on dynamic pricing according to freshness. It relates to the field of online fresh food e-commerce operation technology. The method includes collecting comprehensive product information of target fruits in the online fruit marketplace, continuously monitoring and generating a dynamic state stream containing a time-ordered sequence of remaining inventory and environmental storage parameters. A baseline freshness decay model is determined by matching the batch code and initial freshness assessment value with a freshness decay knowledge base. An adaptive freshness decay prediction model is obtained through real-time environmental storage parameter correction, outputting a freshness prediction trajectory for future time points. Based on the freshness prediction trajectory, the initial points pricing is dynamically reduced to generate a dynamic points pricing list. Combined with the inventory sequence, a collaborative analysis of pricing and inventory, and points redemption matching are completed. This invention can improve the accuracy of freshness prediction and achieve dynamic adaptation of points pricing to product freshness and inventory status.

Description

Technical Field

[0001] This invention belongs to the field of online fresh food e-commerce operation technology, specifically an online fruit mall points redemption method and system based on dynamic pricing based on freshness. Background Technology

[0002] Existing online fruit marketplaces mostly use a fixed points-based pricing mechanism for point redemption, and fruit freshness assessment uses a static, uniform standard. Product inventory and points-based pricing are managed independently. Temperature and humidity environmental parameters during fruit storage are only collected and stored in real time, without being correlated with changes in freshness degradation. Current freshness degradation analysis uses a general, fixed model, without combining it with batch codes and initial freshness assessment values ​​for personalized matching. It also cannot dynamically correct model parameters using real-time environmental storage parameters, and therefore cannot output a predicted freshness trajectory for future time points.

[0003] The current points-based pricing adjustment only considers shelf time and total inventory, failing to take freshness changes as a core adjustment factor, and thus cannot generate a dynamic points-based pricing list that matches future time points. There is no collaborative analysis logic between inventory and pricing data; points redemption only performs basic request matching, failing to reflect the dynamic changes in fruit freshness and inventory. It is necessary to base adjustments on shelf batch codes and initial freshness assessment values, match them with a pre-defined freshness decay knowledge base to determine a baseline freshness decay model, and then use real-time temperature and humidity parameters to perform model calibration, obtaining a freshness prediction trajectory adapted to actual storage conditions. The initial points price needs to be dynamically reduced based on the freshness prediction trajectory to generate a dynamic points-based pricing list corresponding to future time points. This, combined with remaining inventory, should enable collaborative analysis of inventory and pricing, completing a points redemption process that matches the product's status. Summary of the Invention

[0004] This invention aims to solve at least one of the technical problems existing in the prior art; Therefore, this invention proposes a points redemption method for online fruit malls based on dynamic pricing according to freshness, including: Collect comprehensive product information of target fruit products in online fruit malls. The comprehensive product information includes product name, batch code, initial freshness assessment value, initial inventory quantity and initial standard points pricing. The system continuously monitors the real-time status of the target fruit product and generates a dynamic status stream. The dynamic status stream includes a time-ordered sequence of remaining inventory and a sequence of environmental storage parameters, including temperature and humidity readings. Based on the batch code and initial freshness assessment value, a preset freshness decay knowledge base is queried to determine the benchmark freshness decay model corresponding to the target fruit product. Based on the environmental storage parameter sequence, the baseline freshness decay model is corrected in real time to generate an adaptive freshness decay prediction model. Using the adaptive freshness decay prediction model, the freshness of the target fruit product at future time points is predicted, and the freshness prediction trajectory containing the future time points and the predicted freshness values ​​is output. Based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to future time points; The system combines the dynamic points pricing list and the remaining inventory sequence to perform a collaborative analysis of inventory and pricing, and completes the matching and redemption process based on user points redemption requests.

[0005] Further, based on the batch code and initial freshness assessment value, a preset freshness decay knowledge base is queried to determine the baseline freshness decay model corresponding to the target fruit product, including: In the preset freshness decay knowledge base, a primary index is established based on fruit category, and secondary indexes are established based on different origins and seasons; Using the origin and seasonal information contained in the product name and the batch code, the primary and secondary indexes are traversed to locate the historical decay data set that matches the target fruit product; Extract the historical freshness decay curves under standard environmental parameters from the historical decay data set; The initial freshness assessment value is aligned and calibrated with the initial value of the historical freshness decay curve to obtain the calibrated curve as the benchmark freshness decay model.

[0006] Furthermore, based on the environmental storage parameter sequence, the baseline freshness decay model is subjected to real-time parameter correction to generate an adaptive freshness decay prediction model, including: Set a time correction window, and within the time correction window, obtain the average temperature value and average humidity value of the environmental storage parameter sequence; The preset freshness decay influence table is invoked, and the influence factor of the actual environment on the freshness decay rate is obtained by looking up the table based on the average temperature value and average humidity value. The influencing factors are applied to the decay rate parameter of the baseline freshness decay model, and the decay rate parameter is dynamically adjusted. The adjusted decay rate parameter is substituted into the mathematical expression of the baseline freshness decay model to form the adaptive freshness decay prediction model.

[0007] Furthermore, based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to future time points, including: Define a series of future pricing calculation points; Extract the predicted freshness value that corresponds one-to-one with the pricing calculation time point from the freshness prediction trajectory; The freshness retention rate is obtained by calculating the ratio of the predicted freshness value to the initial freshness assessment value at each pricing calculation point. Multiply the initial standard integral pricing by the freshness retention rate to obtain the temporary dynamic integral pricing at the pricing calculation time point; A preset minimum points pricing threshold is introduced. All calculated temporary dynamic points pricing is compared with the minimum points pricing threshold. Temporary dynamic points pricing that is lower than the minimum points pricing threshold is set to the minimum points pricing threshold, thus forming the final dynamic points pricing list.

[0008] Furthermore, a collaborative analysis of inventory and pricing is performed by combining the dynamic points pricing list and the remaining inventory sequence, and a matching and redemption process is completed based on the user's points redemption request, including: By combining the dynamic points-based pricing list and the remaining inventory sequence, a collaborative analysis of inventory and pricing is performed to generate an inventory depletion guidance strategy for the target fruit product. Receive points redemption requests submitted by user accounts, which include the target redemption points value and the desired fruit category; The points redemption request is matched with the dynamic points pricing list at the current moment, and the list of candidate items available for redemption and the actual points deducted for redemption are determined according to the inventory consumption guidance strategy. The candidate product list is returned to the user account, and after the redemption is confirmed, the actual redemption deduction points are deducted from the user account to complete the redemption process; The step involves combining the dynamic points-based pricing list and the remaining inventory sequence to perform a collaborative analysis of inventory and pricing, generating an inventory depletion guidance strategy for the target fruit product, including: Obtain the pricing at each time point in the dynamic points pricing list at the current moment, and the predicted inventory at the corresponding future time points in the remaining inventory sequence; Construct a correlation matrix between inventory consumption and dynamic points pricing. The rows of the correlation matrix represent different inventory consumption rate assumptions, the columns represent different future time points, and the matrix elements are the expected total points consumption calculated based on the pricing list under the given inventory consumption rate and the given future time points. Based on the overall inventory turnover target of the online fruit mall, a range of expected inventory consumption rates is set for the target fruit products. In the correlation matrix, combinations with relatively high expected total consumption of points and corresponding inventory consumption rates within the expected inventory consumption rate range are selected. The pricing time points and suggested redemption amounts corresponding to the selected combinations are encoded as strategy instructions to form the inventory consumption guidance strategy.

[0009] Furthermore, receiving a points redemption request submitted by a user account, which includes the target redemption points value and the desired fruit category, includes: Receive the user's manually entered desired fruit category text from the user's account's profile interface; Receive the target redemption points value manually set by the user from the user's personal center interface; The desired fruit category text is semantically parsed and mapped to a standardized fruit category code, which constitutes the core content of the points redemption request.

[0010] Further, the points redemption request is matched with the current dynamic points pricing list, and based on the inventory consumption guidance strategy, a list of candidate redeemable items and the actual points deducted for redemption are determined, including: Based on the desired fruit category in the points redemption request, all fruit products belonging to the desired fruit category and currently on sale in the online fruit mall are selected as preliminary candidate products. The dynamic points pricing of the preliminary candidate products in the dynamic points pricing list at the current moment is retrieved and compared with the target redemption points value in the points redemption request. Products with dynamic points pricing not higher than the target redemption points value are selected as price candidate products. Read the inventory consumption guidance strategy, and from the price candidate products, prioritize the products recommended for redemption in the strategy to form the final output list of candidate products; When the candidate product list contains multiple products, the actual redemption deduction points value corresponding to each product is calculated according to the priority of the inventory consumption guidance strategy. The actual redemption deduction points value is not higher than its current dynamic points pricing and is not higher than the target redemption points value.

[0011] Furthermore, based on the priority of the aforementioned inventory consumption guidance strategy, the actual redemption deduction points value for each product is calculated, including: Analyze the inventory consumption guidance strategy to obtain the priority weight coefficients set for different products or different redemption time points in the strategy; For each product in the candidate product list, obtain its current dynamic points-based pricing. Input the target redemption points value, the dynamic points pricing of the product, and the priority weight coefficient of the product into the deduction calculation model; The deduction calculation model outputs a suggested actual redemption deduction value based on preset rules. The actual redemption deduction value is usually the dynamic points price of the product. However, when the dynamic points price is too high or the priority weight coefficient is low, the model outputs a value that is lower than the dynamic points price but does not exceed the target redemption value to encourage redemption.

[0012] The deduction calculation model outputs a suggested actual redemption deduction value based on preset rules, including: The core judgment condition of the preset rule is defined as whether the dynamic points pricing of the product exceeds a specific proportion of its initial standard points pricing. For products whose dynamic points pricing does not exceed the specified percentage, the suggested actual redemption points deduction value is directly set to its current dynamic points pricing. For products whose dynamic points pricing exceeds the specified percentage, a discount is calculated based on their priority weighting coefficient. This discount is then subtracted from the current dynamic points pricing to obtain the suggested actual redemption point deduction value. The result is ensured to be greater than or equal to the minimum points pricing threshold and less than or equal to the target redemption point value.

[0013] Furthermore, after collecting comprehensive product information of the target fruit products in the online fruit marketplace, the process also includes: The source credibility of the initial freshness assessment value is verified, including whether the equipment from which the assessment data is sourced has been calibrated and whether the data collection time is within the validity period. The verified initial freshness assessment value is standardized and mapped to a preset unified freshness dimension to form the initial freshness assessment value that can be directly used in subsequent calculations.

[0014] Furthermore, the present invention also includes an online fruit mall points redemption system based on dynamic pricing based on freshness. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the online fruit mall points redemption method based on dynamic pricing based on freshness as described above.

[0015] Compared with the prior art, the beneficial effects of the present invention are: The system queries a pre-defined freshness decay knowledge base using the batch code and initial freshness assessment value to determine the baseline freshness decay model for the target fruit product. Real-time temperature and humidity environmental storage parameter sequences are then used to correct the baseline freshness decay model, generating an adaptive freshness decay prediction model. This model is then used to predict the freshness of the target fruit product at future points in time, outputting a freshness prediction trajectory that includes the predicted freshness value at each future time. The freshness decay model is compatible with the inherent decay characteristics of the corresponding product batch, and real-time storage environment parameters are integrated into the model correction process. The model's adaptability closely matches the actual storage conditions of the product, and the freshness prediction trajectory fully presents the changes in fruit freshness at different future time points, improving the alignment between the freshness prediction results and the actual freshness status of the product.

[0016] Based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to future time points. This dynamic points pricing list is then combined with the remaining inventory sequence to conduct collaborative analysis of inventory and pricing. The matching and redemption process is completed based on user points redemption requests. Points pricing can be adjusted synchronously with the freshness prediction trajectory, and the dynamic points pricing list can correspond to points redemption standards at different future time points. Inventory status data and points pricing data are linked, and user points redemption requests can be matched based on the collaborative analysis of inventory and pricing. This improves the adaptability of the points redemption process to product freshness and inventory status. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of the online fruit mall points redemption method based on freshness-based dynamic pricing as described in this invention. Figure 2 The flowchart for the baseline freshness decay model; Figure 3 A flowchart for generating a dynamic points-based pricing list; Figure 4 A curve for predicting the decline in fruit freshness; Figure 5 This is a graph showing the relationship between the freshness decay model and environmental parameters. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] See Figure 1This invention discloses a method for redeeming points in an online fruit marketplace based on dynamic pricing according to freshness. The method includes: collecting comprehensive product information of a target fruit product in the online marketplace, including product name, batch code, initial freshness assessment value, initial inventory quantity, and initial standard points pricing. The system continuously monitors the real-time status of the target fruit product, generating a dynamic status stream. This stream includes a time-ordered sequence of remaining inventory and a sequence of environmental storage parameters, including temperature and humidity readings. Based on the batch code and initial freshness assessment value, a pre-defined freshness decay knowledge base is queried to determine the baseline freshness decay model corresponding to the target fruit product. The baseline freshness decay model is then real-time parameter corrected based on the environmental storage parameter sequence to generate an adaptive freshness decay prediction model. Using the adaptive freshness decay prediction model, the freshness of the target fruit product at future time points is predicted, outputting a freshness prediction trajectory containing the predicted freshness value at the future time point. Based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to the future time points. By combining a dynamic points-based pricing list and a remaining inventory sequence, a collaborative analysis of inventory and pricing is performed, and a matching and redemption process is completed based on user points redemption requests.

[0020] In one embodiment of the present invention, see [reference] Figure 2 In a pre-defined freshness decay knowledge base, a primary index is created based on fruit category, and secondary indexes are created based on different origins and seasons. Using the origin and season information contained in the product name and batch code, the primary and secondary indexes are traversed to locate the historical decay data set matching the target fruit product. Historical freshness decay curves under standard environmental parameters are extracted from the historical decay data set. The initial freshness assessment value is aligned and calibrated with the initial value of the historical freshness decay curve to obtain the calibrated curve as the baseline freshness decay model. A time correction window is set, and within this window, the average temperature and average humidity values ​​of the environmental storage parameter sequence are obtained. A pre-defined freshness decay influence table is called, and based on the average temperature and average humidity values, the influence factors of the actual environment on the freshness decay rate are obtained from the table. These influence factors are applied to the decay rate parameters of the baseline freshness decay model, dynamically adjusting the decay rate parameters. The adjusted decay rate parameters are then substituted into the mathematical expression of the baseline freshness decay model to form an adaptive freshness decay prediction model.

[0021] In practice, a pre-defined freshness decay knowledge base is established with a primary index based on fruit category and secondary indexes based on different origins and seasons. For example, the primary index includes category entries such as "apple" and "banana," while the secondary index further divides "apple" into combinations of origin and season, such as "Yantai Autumn" and "Shaanxi Summer." Using the origin and season information contained in the product name and batch code, the primary and secondary indexes are traversed to locate the historical decay data set matching the target fruit product. For example, if the product name is "Red Fuji Apple" and the batch code is "YTPP20231015," the origin is "Yantai" and the season is "Autumn" by parsing the code, thus retrieving the corresponding historical decay data set of "apple-Yantai-Autumn" from the knowledge base. Historical freshness decay curves under standard environmental parameters are extracted from the historical decay data set. These curves store historical freshness measurements at a standard temperature of 20 degrees Celsius and a standard humidity of 60% in time series form. The initial freshness assessment value is aligned and calibrated with the initial value of the historical freshness decay curve to obtain the calibrated curve as the benchmark freshness decay model. For example, if the initial freshness assessment value is 95 points and the initial value of the historical freshness decay curve is 90 points, alignment and calibration are achieved by multiplying the entire historical freshness decay curve by a scaling factor of 95 / 90.

[0022] In some embodiments, the baseline freshness decay model is represented by a mathematical expression to support subsequent parameter correction calculations. This expression is defined as:

[0023] in: Indicates a time point under standard environmental parameters. The baseline freshness prediction value, This represents the initial freshness assessment value after alignment and calibration. This represents the baseline decay rate parameter obtained by fitting the historical freshness decay curve. This represents the time offset calculated from the moment the item is put on the shelves. This represents the natural exponential function. A time correction window is set. Within this window, the average temperature and average humidity values ​​of the environmental storage parameter sequence are obtained. For example, if the time correction window length is the most recent hour, the arithmetic mean of all temperature readings within this window is calculated as the average temperature value, and the arithmetic mean of all humidity readings is calculated as the average humidity value. A preset freshness decay influence table is called. Based on the average temperature and average humidity values, the influence factor of the actual environment on the freshness decay rate is obtained from the table. For example, when the average temperature is 25 degrees Celsius and the average humidity is 70%, the influence factor is 1.2. The influence factor is applied to the decay rate parameters of the baseline freshness decay model, and the decay rate parameters are dynamically adjusted. For example, the baseline decay rate parameters... Multiplying by the influence factor 1.2 yields the adjusted decay rate parameter. Substituting the adjusted decay rate parameter into the mathematical expression of the baseline freshness decay model, an adaptive freshness decay prediction model is formed, and its expression is updated as follows:

[0024] in: Indicates the time point under the measured environmental parameters. The adaptive freshness prediction value is provided. Optionally, the length of the time correction window can be configured according to the characteristics of the fruit category. For example, for perishable fruits like strawberries, the time correction window is set to 30 minutes, and for storable fruits like apples, the time correction window is set to 2 hours. In specific implementation, the freshness decay influence table is stored in the form of a two-dimensional matrix. The row index of the matrix corresponds to the discrete temperature range, the column index corresponds to the discrete humidity range, and the matrix elements are predefined influence factor values. These values ​​are obtained through experimental calibration based on fruit and vegetable preservation research data. For example, the temperature range is divided into 0-5 degrees Celsius, 5-10 degrees Celsius, etc., and the humidity range is divided into 50%-60%, 60%-70%, etc., with an influence factor stored at the intersection of each range. It can be understood that when the influence factor acts on the baseline decay rate parameter, a multiplicative operation is used to achieve linear correction, reflecting the accelerating or slowing effect of temperature and humidity on the freshness decay rate.

[0025] In some embodiments, the alignment calibration process can be achieved through linear transformation or nonlinear mapping. For example, when there is a nonlinear relationship between the historical freshness decay curve and the initial freshness assessment value, a piecewise interpolation method is used to proportionally map each point on the historical curve to the curve starting from the initial freshness assessment value. In a specific implementation, the average value of the environmental storage parameter sequence is calculated using a sliding window method. Whenever new monitoring data arrives, the time correction window slides forward by one sampling interval, and the average temperature and average humidity values ​​are recalculated, thereby achieving real-time parameter updates. It can be understood that the benchmark freshness decay model can be in the form of a polynomial or empirical piecewise function, in addition to the exponential form, as long as it can characterize the trend of freshness decay over time. Optionally, the historical decay data set of the freshness decay knowledge base comes from the freshness monitoring records of similar products accumulated in the long-term operation of online fruit malls, or from publicly available fruit and vegetable preservation research databases. These data are stored in the knowledge base after being cleaned and normalized.

[0026] In one embodiment of the present invention, see [reference] Figure 3This process defines a series of future pricing calculation time points and extracts the predicted freshness values ​​corresponding to each time point from the freshness prediction trajectory. The ratio of the predicted freshness value to the initial freshness assessment value is calculated for each pricing calculation time point to obtain the freshness retention rate. The initial standard points price is multiplied by the freshness retention rate to obtain the temporary dynamic points price for each pricing calculation time point. A preset minimum points price threshold is introduced, and all calculated temporary dynamic points prices are compared with this threshold. Temporary dynamic points prices below this threshold are set as the minimum points price threshold, forming the final dynamic points price list.

[0027] In practice, a series of future pricing calculation time points are defined. These time points are evenly distributed along a timeline from the current time to the product's preset shelf-life expiration time at preset time intervals. For example, for a fruit product with a shelf life of 72 hours, the pricing calculation time points are defined as 12, 24, 36, 48, 60, and 72 hours after the current time. Predicted freshness values ​​corresponding one-to-one with the pricing calculation time points are extracted from the freshness prediction trajectory. The freshness prediction trajectory is stored in time series form, with each time point associated with a predicted freshness value. The predicted freshness value corresponding to the pricing calculation time point is retrieved from this sequence through time matching. For example, if the pricing calculation time point is 24 hours, the extracted predicted freshness value is 85 points. The ratio of the predicted freshness value corresponding to each pricing calculation time point to the initial freshness assessment value is calculated to obtain the freshness retention rate. For example, if the initial freshness assessment value is 100 points and the predicted freshness value at 24 hours is 85 points, then the calculated freshness retention rate is 85 / 100 = 0.85.

[0028] In some embodiments, the initial standard integral pricing is multiplied by the freshness retention rate to obtain the temporary dynamic integral pricing at the pricing calculation point. This calculation process is described by a mathematical expression as follows:

[0029] in: Indicates the first Each pricing calculation point The calculated temporary dynamic integral pricing, This indicates the initial standard points pricing for the product. Indicates at time The freshness retention rate is the ratio of the predicted freshness value to the initial freshness assessment value. A preset minimum points pricing threshold is introduced. All calculated temporary dynamic points prices are compared with this threshold. Temporary dynamic points prices below the minimum threshold are set to the minimum threshold, forming the final dynamic points pricing list. For example, if the initial standard points price is 500 points, and the freshness retention rate at 60 hours is 0.3, the calculated temporary dynamic points price is 150 points. Since the system's preset minimum points pricing threshold is 200 points, and 150 points is lower than 200 points, the final dynamic points price at that time is set to 200 points. The dynamic points pricing list is stored in list form, with each entry containing the pricing calculation time point and the corresponding final dynamic points price. Optionally, the pricing calculation time point can be defined at non-uniform intervals, such as increasing the distribution of time points near the end of the product's shelf life and decreasing the distribution of time points in the early stages of product availability. In practical implementation, the calculation of freshness retention rate can introduce a non-linear mapping relationship. For example, when the predicted freshness value is below a certain threshold, the freshness retention rate is calculated using a piecewise function, rather than a simple linear ratio. The minimum points-based pricing threshold can be a globally fixed value or a different value set for different fruit categories. For example, the minimum points-based pricing threshold for strawberries could be set at 150 points, and for apples at 100 points. It can be understood that the calculation formula for temporary dynamic points-based pricing reflects the core logic that points-based pricing decreases linearly with freshness decay.

[0030] In some embodiments, the freshness prediction trajectory is obtained by querying an adaptive freshness decay prediction model. The time offset of the pricing calculation point is input into the model, and the model outputs the corresponding predicted freshness value, which is used as input for calculating the freshness retention rate. In a specific implementation, the dynamic points-based pricing list is stored in memory as a key-value pair data structure or a time-series database. The key is the timestamp of the pricing calculation point, and the value is the corresponding final dynamic points-based pricing value. It can be understood that the introduction of a minimum points-based pricing threshold prevents the dynamic points-based pricing from dropping to an unreasonably low point when freshness is too low, maintaining the stability of the points-based pricing system and the basic value of the product. Optionally, the freshness retention rate can also be adjusted by introducing a compensation coefficient. For example, for a specific promotional product, the freshness retention rate is multiplied by a coefficient greater than 1 when calculating the temporary dynamic points-based pricing to achieve differentiated promotional incentives based on freshness.

[0031] In one embodiment of the invention, a collaborative analysis of inventory and pricing is performed by combining a dynamic points-based pricing list and a remaining inventory sequence to generate an inventory consumption guidance strategy for a target fruit product. The pricing at each time point in the current dynamic points-based pricing list and the predicted inventory at the corresponding future time points in the remaining inventory sequence are obtained. A correlation matrix between inventory consumption and dynamic points-based pricing is constructed. Rows in the correlation matrix represent different inventory consumption rate assumptions, columns represent different future time points, and matrix elements are the expected total points consumption calculated based on the pricing list under the given inventory consumption rate and future time points. Based on the global inventory turnover target of the online fruit mall, an expected inventory consumption rate range is set for the target fruit product. In the correlation matrix, combinations with relatively high expected total points consumption and corresponding inventory consumption rates within the expected inventory consumption rate range are selected. The pricing time points and suggested redemption quantities corresponding to the selected combinations are encoded as strategy instructions to form an inventory consumption guidance strategy. A points redemption request containing the target redemption points value and the desired fruit category is received from a user account. The points redemption request is matched with the current dynamic points pricing list, and based on the inventory depletion guidance strategy, a list of available candidate items and the actual points deducted for redemption are determined. The candidate item list is returned to the user's account, and upon confirmation of redemption, the actual points deducted for redemption are deducted from the user's account, completing the redemption process.

[0032] In practical implementation, a collaborative analysis of inventory and pricing is conducted by combining a dynamic points-based pricing list and a remaining inventory sequence. First, the pricing at each time point in the current dynamic points-based pricing list is obtained, along with the predicted inventory levels at the corresponding future time points in the remaining inventory sequence. For example, the dynamic points-based pricing list might include a price of 450 points for the next 12 hours, 400 points for the next 24 hours, and 350 points for the next 36 hours. The predicted inventory levels at the corresponding time points in the remaining inventory sequence are 80, 50, and 20 units, respectively. A correlation matrix between inventory consumption and dynamic points-based pricing is constructed. The rows of the correlation matrix represent different inventory consumption rate assumptions, the columns represent different future time points, and the matrix elements are the total expected points consumption calculated based on the pricing list under the inventory consumption rate assumptions and the future time points. The formula for the total expected points consumption is:

[0033] in: This indicates the assumption of the inventory consumption rate. and future time points The estimated total value of points consumed below, This indicates the time point in the dynamic points-based pricing list. The pricing, This indicates the assumption of the inventory consumption rate. Down to the time point The estimated quantity of goods consumed at any given time. Inventory consumption rates are assumed to be "slow" (10 items expected to be sold in the next 36 hours), "medium" (30 items), and "fast" (50 items). See Table 1 for an illustration of the correlation matrix calculation. Table 1: Correlation Matrix between Inventory Consumption and Dynamic Incentive Pricing

[0034] Based on the online fruit mall's overall inventory turnover target, a range of expected inventory consumption rates is set for target fruit products. For example, if the target is to achieve a moderately fast turnover, the expected inventory consumption rate range is set to include both "medium speed" and "fast speed" rates. In the correlation matrix, combinations with relatively high expected total points consumption and corresponding inventory consumption rates within the expected inventory consumption rate range are selected. For example, combinations with high expected total points consumption are selected from the table, including (medium speed, 20,000 points in the next 24 hours) and (fast speed, 30,000 points in the next 24 hours). The pricing time points and suggested redemption quantities corresponding to the selected combinations are encoded as strategy instructions to form an inventory consumption guidance strategy. A strategy instruction might be, for example, "In the next 24 hours, it is recommended to guide the redemption of at least 30 items."

[0035] In some embodiments, a points redemption request containing a target redemption point value and a desired fruit category is received from a user account. The user enters the target redemption point value of 800 points and the desired fruit category "apple" through the interface. The points redemption request is matched with the current dynamic points pricing list, and a list of candidate redeemed items and the actual points deduction value are determined according to the inventory consumption guidance strategy. The candidate item list is returned to the user account, and after the user confirms the specific item to be redeemed from the candidate item list, the actual points deduction value is deducted from the user account, completing the redemption process. Optionally, when constructing the correlation matrix, the inventory consumption rate assumption can be simulated based on historical sales data, for example, by analyzing the sales speed of similar products in the same period over the past week, defining three levels of assumptions: "slow," "medium," and "fast." Expected product consumption quantity. Based on inventory consumption rate assumptions and calculations at future time points, for example, under the "medium speed" assumption, 1.25 items are consumed per hour, accumulating to 30 items consumed in the next 24 hours. It can be understood that the row and column definitions of the correlation matrix can be adjusted according to management granularity; for example, the inventory consumption rate assumption can be subdivided into five levels, and future time points can be divided into 6-hour intervals. In some embodiments, the expected inventory consumption rate range is dynamically calculated by the mall operation system based on factors such as overall inventory levels and seasonal demand. In specific implementation, the rule for selecting combinations can be defined as finding all combinations whose expected total points consumption value exceeds a certain threshold, and then selecting combinations whose inventory consumption rate falls within the expected range. It can be understood that the encoding format of the strategy instructions can be a structured data object, containing information such as the suggested time window, suggested redemption item identifier, suggested redemption quantity, and priority weight. Optionally, the calculation of the expected total points consumption value can be corrected by incorporating a discount factor or promotion coefficient to reflect the impact of promotional activities on the actual consumption value of points.

[0036] In one embodiment of the present invention, the system receives the user's manually inputted desired fruit category text and the user's manually set target redemption points value from the user's personal center interface. The desired fruit category text is semantically parsed and mapped to a standardized fruit category code, forming the core content of the points redemption request. Based on the desired fruit category in the points redemption request, all fruit products belonging to the desired fruit category and currently on sale in the online fruit mall are selected as preliminary candidate products. The dynamic points pricing of the preliminary candidate products in the dynamic points pricing list is retrieved at the current moment and compared with the target redemption points value in the points redemption request. Products with dynamic points pricing not higher than the target redemption points value are selected as price candidate products. The inventory consumption guidance strategy is read, and products recommended for redemption in the strategy are prioritized from the price candidate products to form the final output candidate product list. When the candidate product list contains multiple products, the actual redemption deduction points value corresponding to each product is calculated according to the priority of the inventory consumption guidance strategy. The actual redemption deduction points value is not higher than its current dynamic points pricing and not higher than the target redemption points value. The inventory consumption guidance strategy is analyzed to obtain the priority weight coefficients set for different products or different redemption times. For each product in the candidate product list, its current dynamic points price is obtained. The target redemption points value, the product's dynamic points price, and the product's corresponding priority weight coefficient are input into the deduction calculation model. Based on preset rules, the deduction calculation model outputs a suggested actual redemption points deduction value. The actual redemption points deduction value is usually the product's dynamic points price, but when its dynamic points price is too high or its priority weight coefficient is low, the model outputs a value lower than the dynamic points price but not exceeding the target redemption points value to encourage redemption. The core judgment condition of the preset rules is defined: whether the product's dynamic points price exceeds a specific proportion of its initial standard points price. For products whose dynamic points price does not exceed the specific proportion, their suggested actual redemption points deduction value is directly set to their current dynamic points price. For products whose dynamic points price exceeds the specific proportion, a discount is calculated based on their priority weight coefficient. This discount is subtracted from the current dynamic points price to obtain the suggested actual redemption points deduction value, ensuring that the result is greater than or equal to the minimum points price threshold and less than or equal to the target redemption points value.

[0037] In practice, the system receives the desired fruit category text manually entered by the user from their personal account's interface. For example, the user might enter "Red Fuji apple" in the text input box. The system also receives the target redemption points value manually set by the user, such as 800 points in the numeric input box. The desired fruit category text "Red Fuji apple" is semantically parsed and mapped to a standardized fruit category code, for example, the internal standard code "APPLE_RED_FUJI," forming the core content of the points redemption request.

[0038] In some embodiments, based on the desired fruit category code "APPLE_RED_FUJI" in the points redemption request, all fruit products belonging to the "Apple" category and subcategory "Red Fuji" that are currently on sale in the online fruit mall are selected as preliminary candidate products. These preliminary candidate products may include multiple Red Fuji apple products from different batches and origins. The dynamic points pricing of the preliminary candidate products is retrieved from the dynamic points pricing list at the current moment. Assuming the current dynamic points pricing of three preliminary candidate products A, B, and C are 780 points, 820 points, and 750 points respectively, these three prices are compared with the target redemption points value of 800 points in the points redemption request. Products with dynamic points pricing not exceeding 800 points are selected. Therefore, products A (780 points) and C (750 points) meet the criteria and are selected as price candidate products, while product B (820 points) is excluded because its price exceeds the target redemption points value. Read the inventory consumption guidance strategy. Assuming the strategy recommends prioritizing the redemption of product C, select product C from the price candidate products. Product A may also be included. This will form the final output candidate product list, with product C as the first item.

[0039] Optionally, when the candidate product list contains multiple products, the actual redemption deduction points value for each product is calculated based on the priority of the inventory consumption guidance strategy. The actual redemption deduction points value is no higher than its current dynamic points price and no higher than the target redemption points value. The inventory consumption guidance strategy is analyzed to obtain the priority weight coefficients set for different products or different redemption times in the strategy. For example, in the strategy instruction, the priority weight coefficient for product C is 0.9, and the priority weight coefficient for product A is 0.5. For each product in the candidate product list, its current dynamic points price is obtained. The dynamic points price for product C is 750 points, and the dynamic points price for product A is 780 points. The target redemption points value of 800 points, the product's dynamic points price, and the product's corresponding priority weight coefficient are input into the deduction calculation model. The deduction calculation model outputs a suggested actual redemption deduction points value according to preset rules. The actual redemption deduction points value is usually the product's dynamic points price, but when its dynamic points price is too high or its priority weight coefficient is low, the model outputs a value lower than the dynamic points price but not exceeding the target redemption points value to encourage redemption.

[0040] In implementation, the core judgment condition of the preset rule is defined as whether the dynamic points price of a product exceeds a specific percentage of its initial standard points price, for example, 80%. For product C, assuming its initial standard points price is 900 points, the current dynamic points price of 750 points accounts for 83.3% of the initial standard points price, exceeding 80%, thus triggering a branch that calculates the discount based on the priority weight coefficient. For product A, assuming its initial standard points price is 1000 points, the current dynamic points price of 780 points accounts for 78% of the initial standard points price, not exceeding 80%, therefore its suggested actual redemption deduction value is directly set to its current dynamic points price of 780 points. For product C, considering its priority weight coefficient of 0.9, the discount formula is:

[0041] in: This represents the calculated reduction. This indicates the current dynamic points-based pricing for the product (750 points). This indicates the system's preset minimum points pricing threshold (e.g., 200 points). This represents the priority weighting coefficient of the product (0.9). The resulting discount is calculated. Points. The suggested actual points deduction value is obtained by subtracting the discount from the current dynamic points pricing. Points: This result must be greater than or equal to the minimum points pricing threshold of 200 points and less than or equal to the target redemption points value of 800 points. For candidate items and deduction calculations, please refer to Table 2. Table 2: Calculation Table of Actual Redemption Points for Candidate Goods

[0042] It is understandable that the preset rules of the deduction calculation model can set different specific ratio thresholds to adjust the triggering conditions of dynamic pricing deductions. In some embodiments, the priority weight coefficient is dynamically assigned by the inventory consumption guidance strategy based on the degree of urgent consumption of the product, with a value range between 0 and 1. The higher the weight coefficient, the more the strategy encourages the redemption of the product. In specific implementations, the formula for calculating the deduction amount can be designed in other forms, such as the deduction amount being proportional to the product of (current dynamic points pricing - minimum points pricing threshold) and (1 - priority weight coefficient), to ensure that high-priority products have lower deductions and points values ​​closer to their dynamic pricing, while low-priority products have higher deductions and points values ​​to enhance their attractiveness. It is understandable that when the dynamic points pricing of a product does not exceed the core judgment conditions, its dynamic points pricing is directly used as the actual deduction value, simplifying the calculation and respecting the established pricing. Optionally, the final list of candidate products displayed to the user will be sorted from low to high based on the actual redemption points value or by strategy priority, and the actual redemption points value of each product will be clearly displayed for the user to choose from.

[0043] See Figure 4 This is a fruit freshness decay prediction curve, visually illustrating the freshness decay patterns of two types of Fuji apples (Product A and Product C) over a 10-day storage period. By analyzing the differences in decay rates, it provides a basis for inventory consumption guidance strategies and can be used to verify the accuracy of the model after environmental parameter correction, comparing the decay differences between standard and actual storage conditions. It intuitively presents the calculation results of the baseline freshness decay model and the environmental correction model in the patent, transforming abstract mathematical formulas into perceptible trend curves, clearly demonstrating the differentiated decay characteristics of different batches of fruit, and providing a data foundation for subsequent dynamic pricing. It can be directly used in the e-commerce operation backend to monitor the freshness status of products in real time, assisting operational decisions, conforming to the physical laws of fruit freshness decay, and facilitating subsequent prediction and modeling.

[0044] In one embodiment of the present invention, after collecting full-dimensional product information of the target fruit product in an online fruit marketplace, the initial freshness assessment value is verified for source credibility. This verification includes verifying whether the data source device has been calibrated and whether the data collection time is within the validity period. The verified initial freshness assessment value is then standardized and mapped to a preset unified freshness dimension to form an initial freshness assessment value that can be directly used in subsequent calculations.

[0045] In practice, after collecting comprehensive product information on the target fruit from the online fruit marketplace, the initial freshness assessment value undergoes source credibility verification. This verification process includes assessing whether the data source device has been calibrated and whether the data collection time is within its validity period. For example, if the initial freshness assessment value comes from an IoT spectral detection device, the system queries the device's calibration records and confirms that it has been calibrated within the past 7 days and has a valid calibration certificate; in this case, the device calibration status is deemed successful. Simultaneously, the system reads the data collection timestamp attached to the initial freshness assessment value and compares it with the current system time. If the time difference is less than a preset 2-hour validity threshold, the data collection time is deemed valid. Only when both the device calibration status and the validity of the data collection time pass verification is the initial freshness assessment value considered reliable and proceeds to subsequent processing steps.

[0046] In some embodiments, the validated initial freshness assessment values ​​are standardized and mapped to a preset unified freshness dimension. The preset unified freshness dimension is a numerical range from 0 to 100, where 100 represents optimal freshness and 0 represents inedibility. Standardization transforms initial freshness assessment values ​​from different suppliers or testing equipment, with different dimensions and benchmarks, into this unified range. For example, one supplier provides an initial freshness assessment value of "Grade A," corresponding to its internal standard range of 90-100 points, while another supplier provides an initial freshness assessment value of "92," corresponding to its percentage system. By consulting a preset mapping table, "Grade A" is mapped to 95 points under the unified freshness dimension, and the value "92" is directly used as 92 points under the unified freshness dimension. The mapping table records the correspondence between various common freshness representation methods and unified freshness dimension scores.

[0047] Optionally, the standardization process can be implemented using a linear transformation formula, which is expressed as follows:

[0048] in: This represents the standardized initial freshness assessment value mapped to a preset uniform freshness unit. This represents the original initial freshness assessment value after verification. This represents the scaling factor determined based on the original assessment value's dimensions and the standardized freshness measurement dimensions. This represents the offset coefficient determined based on the dimensions of the original evaluation value and the unified freshness dimension. For example, the original evaluation value output by a certain testing device. The unit of measurement is 0-10, while the unit of measurement for uniform freshness is 0-100. A scaling factor is set. Offset coefficient The standardization process involves multiplying the original value by 10 to obtain a score under a unified dimension. In some embodiments, the offset coefficient... This is used to adjust the baseline. For example, when 60 points in a certain evaluation system corresponds to 80 points in a unified freshness measurement system, an appropriate baseline needs to be set. and To achieve this mapping.

[0049] In practical implementation, a pre-set mapping table or linear transformation coefficients are used. and The system is pre-configured based on fruit category and evaluation method during system initialization. It is understood that the calibration verification of data source devices can be extended to checking whether the device serial number is in the whitelist of the marketplace certification, or checking whether the detailed report of the device's most recent calibration has no abnormal alarms. The verified initial freshness assessment values ​​are standardized to form initial freshness assessment values ​​that can be directly used in subsequent calculations, ensuring that freshness data from diverse and heterogeneous data sources have a consistent order of magnitude and comparability before entering the freshness decay model calculation. Optionally, the threshold for the validity period of data collection time can be differentiated according to the perishability of fruit categories; for example, the validity period threshold for strawberries is set to 1 hour, and for apples, it is set to 4 hours. In some embodiments, if the initial freshness assessment value fails the source credibility verification, the system can trigger an alarm and attempt to obtain from an alternate data source or use the default initial freshness assessment value for that category of product for subsequent calculations. It is understandable that the preset range of the unified freshness measurement unit is not limited to 0-100 points, but can also be a decimal of 0-1 or other continuous numerical ranges. The core is to provide a unified and unambiguous numerical benchmark for subsequent freshness decay calculations.

[0050] See Figure 5 This is a graph showing the relationship between a freshness decay model and environmental parameters, fully illustrating the correlation between the baseline model, the adaptive model, and environmental parameters. Temperature and humidity data are the core inputs for adjusting the decay rate parameter, directly affecting the accuracy of the adaptive model. The adaptive model's prediction results are the core data for dynamic point-based pricing, more realistic than the baseline model, avoiding pricing deviations. It can monitor in real time whether the storage environment meets the requirements for fruit preservation, triggering alerts when anomalies occur. Based on the actual decay rate, it adjusts inventory consumption guidance strategies, prioritizing the handling of products that decay faster. It can be directly used in the e-commerce warehouse management backend to monitor product freshness and storage environment in real time, assisting operational decisions by integrating the two dimensions of freshness decay and environmental parameters into a single graph, clearly demonstrating the causal relationship.

[0051] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A points redemption method for online fruit malls based on dynamic pricing according to freshness, characterized in that... include: Collect comprehensive product information of target fruit products in online fruit malls. The comprehensive product information includes product name, batch code, initial freshness assessment value, initial inventory quantity and initial standard points pricing. The system continuously monitors the real-time status of the target fruit product and generates a dynamic status stream. The dynamic status stream includes a time-ordered sequence of remaining inventory and a sequence of environmental storage parameters, including temperature and humidity readings. Based on the batch code and initial freshness assessment value, a preset freshness decay knowledge base is queried to determine the benchmark freshness decay model corresponding to the target fruit product. Based on the environmental storage parameter sequence, the baseline freshness decay model is corrected in real time to generate an adaptive freshness decay prediction model. Using the adaptive freshness decay prediction model, the freshness of the target fruit product at future time points is predicted, and the freshness prediction trajectory containing the future time points and the predicted freshness values ​​is output. Based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to future time points; The system combines the dynamic points pricing list and the remaining inventory sequence to perform a collaborative analysis of inventory and pricing, and completes the matching and redemption process based on user points redemption requests.

2. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 1, characterized in that, Based on the batch code and initial freshness assessment value, a preset freshness decay knowledge base is queried to determine the baseline freshness decay model corresponding to the target fruit product, including: In the preset freshness decay knowledge base, a primary index is established based on fruit category, and secondary indexes are established based on different origins and seasons; Using the origin and seasonal information contained in the product name and the batch code, the primary and secondary indexes are traversed to locate the historical decay data set that matches the target fruit product; Extract the historical freshness decay curves under standard environmental parameters from the historical decay data set; The initial freshness assessment value is aligned and calibrated with the initial value of the historical freshness decay curve to obtain the calibrated curve as the benchmark freshness decay model.

3. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 2, characterized in that, Based on the environmental storage parameter sequence, the baseline freshness decay model is corrected in real time to generate an adaptive freshness decay prediction model, including: Set a time correction window, and within the time correction window, obtain the average temperature value and average humidity value of the environmental storage parameter sequence; The preset freshness decay influence table is invoked, and the influence factor of the actual environment on the freshness decay rate is obtained by looking up the table based on the average temperature value and average humidity value. The influencing factors are applied to the decay rate parameter of the baseline freshness decay model, and the decay rate parameter is dynamically adjusted. The adjusted decay rate parameter is substituted into the mathematical expression of the baseline freshness decay model to form the adaptive freshness decay prediction model.

4. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 3, characterized in that, Based on the freshness prediction trajectory, the initial standard points pricing is dynamically reduced to generate a dynamic points pricing list corresponding to future time points, including: Define a series of future pricing calculation points; Extract the predicted freshness value that corresponds one-to-one with the pricing calculation time point from the freshness prediction trajectory; The freshness retention rate is obtained by calculating the ratio of the predicted freshness value to the initial freshness assessment value at each pricing calculation point. Multiply the initial standard integral pricing by the freshness retention rate to obtain the temporary dynamic integral pricing at the pricing calculation time point; A preset minimum points pricing threshold is introduced. All calculated temporary dynamic points pricing is compared with the minimum points pricing threshold. Temporary dynamic points pricing that is lower than the minimum points pricing threshold is set to the minimum points pricing threshold, thus forming the final dynamic points pricing list.

5. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 4, characterized in that, The system combines the dynamic points pricing list and the remaining inventory sequence to perform collaborative analysis of inventory and pricing, and completes the matching and redemption process based on user points redemption requests, including: By combining the dynamic points-based pricing list and the remaining inventory sequence, a collaborative analysis of inventory and pricing is performed to generate an inventory depletion guidance strategy for the target fruit product. Receive points redemption requests submitted by user accounts, which include the target redemption points value and the desired fruit category; The points redemption request is matched with the dynamic points pricing list at the current moment, and the list of candidate items available for redemption and the actual points deducted for redemption are determined according to the inventory consumption guidance strategy. The candidate product list is returned to the user account, and after the redemption is confirmed, the actual redemption deduction points are deducted from the user account to complete the redemption process; The step involves combining the dynamic points-based pricing list and the remaining inventory sequence to perform a collaborative analysis of inventory and pricing, generating an inventory depletion guidance strategy for the target fruit product, including: Obtain the pricing at each time point in the dynamic points pricing list at the current moment, and the predicted inventory at the corresponding future time points in the remaining inventory sequence; Construct a correlation matrix between inventory consumption and dynamic points pricing. The rows of the correlation matrix represent different inventory consumption rate assumptions, the columns represent different future time points, and the matrix elements are the expected total points consumption calculated based on the pricing list under the given inventory consumption rate and the given future time points. Based on the overall inventory turnover target of the online fruit mall, a range of expected inventory consumption rates is set for the target fruit products. In the correlation matrix, combinations with relatively high expected total consumption of points and corresponding inventory consumption rates within the expected inventory consumption rate range are selected. The pricing time points and suggested redemption amounts corresponding to the selected combinations are encoded as strategy instructions to form the inventory consumption guidance strategy.

6. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 5, characterized in that, The process of receiving a points redemption request submitted by a user account, which includes the target redemption points value and the desired fruit category, includes: Receive the user's manually entered desired fruit category text from the user's account's profile interface; Receive the target redemption points value manually set by the user from the user's personal center interface; The desired fruit category text is semantically parsed and mapped to a standardized fruit category code, which constitutes the core content of the points redemption request.

7. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 6, characterized in that, The points redemption request is matched with the current dynamic points pricing list, and based on the inventory consumption guidance strategy, a list of candidate redeemable items and the actual points deducted for redemption are determined, including: Based on the desired fruit category in the points redemption request, all fruit products belonging to the desired fruit category and currently on sale in the online fruit mall are selected as preliminary candidate products. The dynamic points pricing of the preliminary candidate products in the dynamic points pricing list at the current moment is retrieved and compared with the target redemption points value in the points redemption request. Products with dynamic points pricing not higher than the target redemption points value are selected as price candidate products. Read the inventory consumption guidance strategy, and from the price candidate products, prioritize the products recommended for redemption in the strategy to form the final output list of candidate products; When the candidate product list contains multiple products, the actual redemption deduction points value corresponding to each product is calculated according to the priority of the inventory consumption guidance strategy. The actual redemption deduction points value is not higher than its current dynamic points pricing and is not higher than the target redemption points value.

8. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 7, characterized in that, Based on the priority of the aforementioned inventory consumption guidance strategy, calculate the actual redemption deduction points value for each product, including: Analyze the inventory consumption guidance strategy to obtain the priority weight coefficients set for different products or different redemption time points in the strategy; For each product in the candidate product list, obtain its current dynamic points-based pricing. Input the target redemption points value, the dynamic points pricing of the product, and the priority weight coefficient of the product into the deduction calculation model; The deduction calculation model outputs a suggested actual redemption deduction value based on preset rules. This actual redemption deduction value is typically the product's dynamic points pricing. However, when the dynamic points pricing is too high or the priority weight coefficient is low, the model outputs a value lower than the dynamic points pricing but not exceeding the target redemption point value to encourage redemption. The deduction calculation model outputs a suggested actual redemption deduction value based on preset rules, including: The core judgment condition of the preset rule is defined as whether the dynamic points pricing of the product exceeds a specific proportion of its initial standard points pricing. For products whose dynamic points pricing does not exceed the specified percentage, the suggested actual redemption points deduction value is directly set to its current dynamic points pricing. For products whose dynamic points pricing exceeds the specified percentage, a discount is calculated based on their priority weighting coefficient. This discount is then subtracted from the current dynamic points pricing to obtain the suggested actual redemption point deduction value. The result is ensured to be greater than or equal to the minimum points pricing threshold and less than or equal to the target redemption point value.

9. The online fruit mall points redemption method based on freshness-based dynamic pricing according to claim 8, characterized in that, After collecting comprehensive product information on the target fruit products from the online fruit marketplace, the process also includes: The source credibility of the initial freshness assessment value is verified, including whether the data source device has been calibrated and whether the data collection time is within the validity period. The verified initial freshness assessment value is standardized and mapped to a preset unified freshness dimension to form the initial freshness assessment value that can be directly used in subsequent calculations.

10. An online fruit mall points redemption system based on freshness-based dynamic pricing, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the online fruit mall points redemption method based on freshness dynamic pricing as described in any one of claims 1 to 9.

Images