Method and device for adjusting expected delivery time, electronic equipment and storage medium

Abstract

This application discloses a method, apparatus, electronic device, and storage medium for adjusting estimated delivery time. The method includes: acquiring the estimated delivery time (ETA) of a current order and order-related information; determining a current sensitive scenario corresponding to the current order based on the ETA and the order-related information, wherein the current sensitive scenario is a scenario related to delivery efficiency and order conversion rate; determining the current time range to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and time ranges, wherein the correspondence between the sensitive scenario and time ranges is determined in advance based on historical data; adjusting the ETA according to the current time range; and outputting the adjusted ETA. The current sensitive scenario obtained in this application embodiment is a result that comprehensively considers delivery efficiency and order conversion rate, thereby improving delivery efficiency while ensuring order volume.

Description

Technical Field

[0001] This application relates to the field of Internet technology, and in particular to a method, apparatus, electronic device, and storage medium for adjusting the estimated delivery time. Background Technology

[0002] ETA (Estimated Time of Arrival) is a core parameter in delivery operations. ETA has a significant impact on order size, user experience, and delivery efficiency. From a user perspective, an excessively long ETA leads to a poor user experience and a lower order conversion rate, while an excessively short ETA may result in on-time delivery, negatively affecting the user experience. Regarding delivery efficiency, ETA is a scheduling constraint. An excessively short ETA leaves no room for scheduling strategies, hindering efficiency improvements. Conversely, an excessively long ETA in inappropriate situations (such as low spatiotemporal density, i.e., low order density) wastes delivery efficiency and negatively impacts the user experience.

[0003] In related technologies, ETA primarily relies on model predictions, with the model's target being historical offline delivery times. The goal is accuracy, meaning improving prediction precision. However, because the model prioritizes accuracy, it neglects delivery efficiency, resulting in inadequate delivery efficiency. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, and storage medium for adjusting the estimated delivery time, which helps to improve delivery efficiency.

[0005] To address the aforementioned problems, in a first aspect, embodiments of this application provide a method for adjusting the estimated delivery time, comprising:

[0006] Get the estimated delivery time (ETA) and order-related information for the current order;

[0007] Based on the ETA and the order-related information, determine the current sensitive scenario corresponding to the current order. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate.

[0008] Based on the current sensitive scenario and the correspondence between the sensitive scenario and the time tier, the current time tier to be adjusted for the ETA is determined, wherein the correspondence between the sensitive scenario and the time tier is determined in advance based on historical data.

[0009] The ETA is adjusted according to the current time range, and the adjusted ETA is output.

[0010] Secondly, embodiments of this application provide an adjustment device for the estimated delivery time, comprising:

[0011] The order information acquisition module is configured to acquire the estimated delivery time (ETA) and related order information for the current order.

[0012] The scenario determination module is configured to determine the current sensitive scenario corresponding to the current order based on the ETA and the order-related information. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate.

[0013] The time tier determination module is configured to determine the current time tier to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time tier, wherein the correspondence between the sensitive scenario and the time tier is determined in advance based on historical data.

[0014] The ETA adjustment module is configured to adjust the ETA according to the current time range and output the adjusted ETA.

[0015] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the estimated delivery time adjustment method described in embodiments of this application.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for adjusting the estimated delivery time disclosed in embodiments of this application.

[0017] The estimated delivery time adjustment method, apparatus, electronic device, and storage medium provided in this application embodiment obtain the ETA and order-related information of the current order, determine the current sensitive scenario corresponding to the current order based on the ETA and order-related information, determine the current time range of the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time range, adjust the ETA based on the current time range, and output the adjusted ETA. Since the current sensitive scenario is related to delivery efficiency and order conversion rate, the obtained current sensitive scenario is a result that comprehensively considers delivery efficiency and order conversion rate, thereby improving delivery efficiency while ensuring order volume. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram illustrating the changes in delivery order volume and individual delivery capacity efficiency over time in the embodiments of this application;

[0020] Figure 2 This is a flowchart of a method for adjusting the estimated delivery time according to Embodiment 1 of this application;

[0021] Figure 3 This is a flowchart of a method for adjusting the estimated delivery time according to Embodiment 2 of this application;

[0022] Figure 4a This is a schematic diagram illustrating the change in delivery efficiency with ETA in the embodiments of this application;

[0023] Figure 4b This is a schematic diagram illustrating the change in order conversion rate with ETA in the embodiments of this application.

[0024] Figure 5 This is a schematic diagram of the structure of an estimated delivery time adjustment device according to Embodiment 3 of this application;

[0025] Figure 6 This is a schematic diagram of the structure of an electronic device according to Embodiment 4 of this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] Accuracy has long been the primary goal of model predictions for ETA (Electronic Time Assurance), but this approach has also presented challenges. Simply focusing on accuracy may overlook many issues, and it may not significantly impact business metrics. Meanwhile, businesses need to provide more than just accuracy. Finding a better way to align business objectives with user experience while simultaneously improving efficiency presents a dual challenge for both technology and business.

[0028] Figure 1 This is a schematic diagram illustrating the changes in delivery order volume and individual delivery capacity efficiency over time in embodiments of this application, such as... Figure 1As shown, curve 1 represents the change in the number of delivery orders over time, and curve 2 represents the change in the delivery efficiency of a single delivery capacity over time. It can be seen that the number of delivery orders has increased significantly over time, meaning that food delivery has achieved a significant scale advantage with a substantial increase in the number of orders. However, the delivery efficiency of a single delivery capacity has not increased accordingly; it has only increased slowly and has not significantly transformed the scale advantage into a cost advantage.

[0029] There are many reasons for this, but based on business understanding, ETA has a significant impact on delivery efficiency. Specifically, a longer ETA allows for greater flexibility in order consolidation and package merging, increases the spatiotemporal density within the system, and raises the overall optimization ceiling, thus leading to a substantial improvement in efficiency. Therefore, the lack of corresponding iterations to the ETA may have limited the room for adjusting delivery efficiency. The goal is to further unlock delivery efficiency potential and improve overall delivery efficiency while ensuring user experience and scalability.

[0030] Based on this, the model's predicted results and the rule-based results can be merged, and both output as the final ETA result. Since delivery places great emphasis on fulfillment quality, several time-addition rules can be set to appropriately extend the ETA based on factors such as load, distance, and food preparation, ensuring timely delivery and guaranteeing a positive user experience.

[0031] Based on this, the inventors conducted a small experiment, adjusting delivery efficiency through time-addition rules. Specifically, in a certain city, the ETA was uniformly increased by 5 minutes, resulting in a 2.75 percentage point increase in order spatiotemporal density, and a 3.72 percentage point increase during the midday peak. This left significant room for improving delivery efficiency, verifying that ETA can significantly affect spatiotemporal density, thereby impacting delivery capacity and efficiency. This application's embodiment adjusts the ETA based on model-predicted ETA to improve the platform's overall delivery efficiency while maintaining user experience and scale. The specific solution is as follows.

[0032] Example 1

[0033] This embodiment provides a method for adjusting the estimated delivery time, such as... Figure 2 As shown, the method includes one or more of steps 210 to 240.

[0034] Step 210: Obtain the ETA and order-related information for the current order.

[0035] The ETA (Electronic Delivery Amount) is estimated using a model based on information such as order-related data, delivery time, the number of pending orders around the delivery address, and delivery capacity. This can be called the baseline ETA. To improve delivery efficiency while maintaining user experience and scalability, the ETA can be adjusted slightly; this dynamically adjustable ETA can be called the dynamic ETA. The baseline ETA aims to ensure fairness, reasonableness, stability, and controllability at the national level, while the dynamic ETA considers the ability to adjust the ETA to order size and delivery efficiency by combining scheduling information and external information. The current order includes orders currently awaiting generation or orders already generated.

[0036] The order information for the current order can be obtained from the order submission page in the client. Based on this information, the delivery time period, the number of orders to be delivered in the delivery address area, and the delivery capacity can be determined. The ETA (Earnings To Acquired) can be estimated using a model based on this information, thus obtaining the ETA and related order information for the current order that requires ETA adjustment. The order information includes: merchant information, dish type, dish quantity, delivery address, neighborhood type, and delivery time period.

[0037] Step 220: Based on the ETA and the order-related information, determine the current sensitive scenario corresponding to the current order. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate.

[0038] The sensitive scenarios can be categorized based on their sensitivity to delivery efficiency and order conversion rate. The current sensitive scenario is the one corresponding to the current order. Order conversion rate refers to the probability that an order will be submitted.

[0039] For example, in the food delivery industry, adding 1 minute to the ETA during the lunch rush has virtually no impact on order conversion rates, while adding 1 minute to the ETA between 3-4 pm might discourage users from placing orders. Adding 1 minute to an ETA of 30 minutes can significantly increase the probability of order consolidation, thus significantly improving delivery efficiency, while adding 1 minute to an ETA of 60 minutes may have no impact on delivery efficiency. Therefore, it is necessary to determine the sensitive scenarios corresponding to the current order based on order-related information and the ETA.

[0040] The direct objective of this application embodiment is to improve delivery efficiency while maintaining a stable order volume. However, this direct objective is difficult to quantify. Therefore, this application embodiment establishes a quantitative relationship between indirect indicators, namely, it establishes sensitive scenarios and associates these scenarios with order volume and delivery efficiency, thereby achieving the direct objective. When determining the current sensitive scenario corresponding to the current order, under the ETA (Earning Time of Acquisition), information that can affect delivery efficiency and information that can affect order conversion rate are extracted from the order-related information of the current order. Based on this information, the current sensitive scenario corresponding to the current order is determined.

[0041] In one embodiment of this application, determining the current sensitivity scenario corresponding to the current order based on the ETA and the order-related information includes: determining the delivery efficiency impact factor value and the order conversion rate impact factor value corresponding to the ETA based on the ETA and the order-related information; and determining the current sensitivity scenario corresponding to the current order based on the delivery efficiency impact factor value and the order conversion rate impact factor value.

[0042] Through data analysis, different influencing factors are selected to identify those that simultaneously affect ETA and order conversion rate, serving as order conversion rate influencing factors. Similarly, factors that simultaneously affect ETA and delivery efficiency are identified as delivery efficiency influencing factors. In the food delivery sector, order conversion rate influencing factors may include at least one of the following: order pickup difficulty, order delivery difficulty, and time period. Delivery efficiency influencing factors may include at least one of the following: regional load, regional order quantity, and time period. Regional load is the ratio of total order quantity to total delivery capacity; regional order quantity is the number of orders within a certain delivery address range; and total delivery capacity is the number of delivery vehicles responsible for deliveries within that range. Order pickup difficulty can be determined by a combination of factors such as merchant information, dish type, and dish quantity; order delivery difficulty can be determined by a combination of factors such as user address and neighborhood type.

[0043] After obtaining the order-related information for the current order, the delivery efficiency influencing factor value and the order conversion rate influencing factor value corresponding to the ETA are determined from the order-related information. For example, the values ​​of influencing factors such as regional load, regional order quantity, and time period in the delivery efficiency influencing factor are determined, and the values ​​of influencing factors such as order pickup difficulty, order delivery difficulty, and time period in the order conversion rate influencing factor are determined. Taking into account the values ​​of each delivery efficiency influencing factor, it is determined whether the current order is sensitive to delivery efficiency; taking into account the values ​​of each order conversion rate influencing factor, it is determined whether the current order is sensitive to order conversion rate; based on the results of the determination of whether the current order is sensitive to delivery efficiency and whether the current order is sensitive to order conversion rate, the current sensitivity scenario corresponding to the current order is determined.

[0044] The judgment results regarding sensitivity to delivery efficiency can include three categories: highly sensitive to delivery efficiency, moderately sensitive to delivery efficiency, and insensitive to delivery efficiency. Similarly, the judgment results regarding sensitivity to order conversion rate can include three categories: highly sensitive to order conversion rate, moderately sensitive to order conversion rate, and insensitive to order conversion rate. Therefore, combining the three judgment results for delivery efficiency and the three judgment results for order conversion rate, nine sensitivity scenarios can be derived: highly sensitive to both delivery efficiency and order conversion rate, moderately sensitive to both delivery efficiency and order conversion rate, insensitive to both delivery efficiency and order conversion rate, highly sensitive to both delivery efficiency and order conversion rate, moderately sensitive to both delivery efficiency and order conversion rate, insensitive to both delivery efficiency and order conversion rate, and insensitive to both delivery efficiency and order conversion rate. The above-mentioned judgment results on whether delivery efficiency is sensitive and whether order conversion rate is sensitive are just examples. In addition to the three judgment results mentioned above, there are many more types of judgment results on whether delivery efficiency is sensitive and whether order conversion rate is sensitive, resulting in more sensitive scenarios.

[0045] For example, if the current order is determined to be highly sensitive to delivery efficiency based on the delivery efficiency impact factor value, and insensitive to order conversion rate based on the order conversion rate impact factor value, then the current sensitivity scenario for the current order can be determined to be highly sensitive to delivery efficiency but insensitive to order conversion rate.

[0046] In one embodiment of this application, determining the current sensitive scenario corresponding to the current order based on the delivery efficiency impact factor value and the order conversion rate impact factor value includes: determining the current sensitive scenario corresponding to the current order based on the relationship between the delivery efficiency impact factor value and the threshold corresponding to the delivery efficiency impact factor, and the relationship between the order conversion rate impact factor value and the threshold corresponding to the order conversion rate impact factor.

[0047] Each influencing factor corresponds to a threshold. After determining the values ​​of the delivery efficiency influencing factor and the order conversion rate influencing factor, the relationship between the delivery efficiency influencing factor value and its corresponding threshold can be compared to determine whether the current order is sensitive to delivery efficiency. Similarly, the relationship between the order conversion rate influencing factor value and its threshold can be compared to determine whether the current order is sensitive to order conversion rate. Based on these assessments, the current sensitivity scenario for the current order can be determined. For example, the order conversion rate influencing factors include order pickup difficulty, order delivery difficulty, and time period. The current order has an order pickup difficulty of 0.8, an order delivery difficulty of 0.7, and a time period of 11:00-12:00. The thresholds for both order pickup difficulty and order delivery difficulty are 0.6, and the time period threshold is also 11:00-12:00. Therefore, by comparing the values ​​of each order conversion rate influencing factor with their respective thresholds, it can be determined that the current order is relatively sensitive to order conversion rate.

[0048] By comparing the value of the delivery efficiency impact factor with the threshold of the delivery efficiency impact factor, and comparing the value of the order conversion rate impact factor with the threshold of the order conversion rate impact factor, the current sensitivity scenario corresponding to the current order can be determined relatively accurately and quickly.

[0049] Step 230: Determine the current time frame to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time frame. The correspondence between the sensitive scenario and the time frame is determined in advance based on historical data.

[0050] Offline methods determine the correspondence between sensitive scenarios and time intervals based on historical data. A bandit mechanism can be used to determine the optimal time interval for each sensitive scenario, aiming to improve delivery efficiency while maintaining a stable order volume. For example, if the current sensitive scenario is sensitive to delivery efficiency but not to order conversion rate, time can be increased beyond the ETA to improve delivery efficiency.

[0051] The time intervals can be divided into minute increments, such as +5min, -5min, +4min, -4min, +3min, -3min, +2min, -2min, +1min, -1min, etc. By dividing the time intervals into minute increments, the optimization space can be greatly reduced, requiring only the selection of one time interval from a large number of intervals given the ETA and the current sensitivity scenario.

[0052] After determining the current sensitive scenario corresponding to the current order, the current sensitive scenario is matched with the sensitive scenarios in the pre-determined correspondence between sensitive scenarios and time slots. The sensitive scenario that matches the current sensitive scenario in the correspondence is determined, and the time slot corresponding to the sensitive scenario is taken as the current time slot of the ETA.

[0053] Step 240: Adjust the ETA according to the current time range and output the adjusted ETA.

[0054] After determining the current time bracket of the ETA, add the current time bracket to the ETA to adjust it accordingly. For example, if the current time bracket is -3 minutes and the ETA is 30 minutes, add 30 minutes to -3 minutes to obtain an adjusted ETA of 27 minutes. After adjusting the ETA, output the adjusted ETA, for example, by displaying it on the user's order information display page.

[0055] In one embodiment of this application, outputting the adjusted ETA includes: outputting the adjusted ETA to a client for display on the client's order submission page or order details page. The order submission page is the page where an order is submitted, and may include information such as delivery address, order product information, and merchant information. The order details page is the page that displays detailed order information or delivery information after the user submits the order. By displaying the adjusted ETA on the client's order submission page or order details page, users can understand the estimated delivery time of the goods.

[0056] The estimated delivery time adjustment method provided in this application embodiment obtains the ETA and order-related information of the current order, determines the current sensitive scenario corresponding to the current order based on the ETA and order-related information, determines the current time range of the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time range, adjusts the ETA based on the current time range, and outputs the adjusted ETA. Since the current sensitive scenario is related to delivery efficiency and order conversion rate, the obtained current sensitive scenario is a result that comprehensively considers delivery efficiency and order conversion rate, thereby improving delivery efficiency while ensuring order volume.

[0057] Based on the above technical solution, the method may also optionally include: determining the uncontrollable factors corresponding to the current order;

[0058] The step of determining the current sensitivity scenario corresponding to the current order based on the ETA and the order-related information includes: determining the current sensitivity scenario corresponding to the current order based on the ETA, the order-related information, and uncontrollable factors.

[0059] Uncontrollable factors can be uncontrollable elements, such as sudden increases or decreases in order volume, or severe weather, which can affect the choice of delivery time slots. By adding uncontrollable factors to the modeling of sensitive scenarios, we can obtain sensitive scenarios related to delivery efficiency, order conversion rate, ETA, and uncontrollable factors.

[0060] When determining the current sensitivity scenario for a given order, it's also necessary to identify the uncontrollable factors associated with it. This determination is based on the order's ETA, order-related information, and uncontrollable factors. Specifically, it involves determining the delivery efficiency impact factor and order conversion rate impact factor values ​​based on the order's relevant information. Furthermore, it considers the relationships between the delivery efficiency impact factor values ​​and their corresponding thresholds, the order conversion rate impact factor values ​​and their corresponding thresholds, the ETA, and the uncontrollable factors to determine the current sensitivity scenario. By taking uncontrollable factors into account, the decision-making information required for each time period can be further optimized, leading to more comprehensive decisions and ultimately improving delivery efficiency while maintaining order volume.

[0061] Example 2

[0062] This embodiment provides a method for adjusting the estimated delivery time, such as... Figure 3 As shown, the method includes one or more of steps 310 to 370.

[0063] Step 310: Based on the order-related information of each historical order in the historical data, determine the sensitive scenario corresponding to each historical order.

[0064] The correspondence between sensitive scenarios and time periods can be optimized based on historical data. First, the sensitive scenario corresponding to each historical order is determined based on the order-related information of each historical order in the historical data. The method for determining the sensitive scenario corresponding to each historical order is the same as the method for determining the current sensitive scenario corresponding to the current order based on the order-related information of the current order in the above embodiment, and will not be repeated here.

[0065] Step 320: Determine the historical time frame corresponding to each historical order based on the sensitive scenario corresponding to each historical order.

[0066] Based on the sensitive scenarios corresponding to each historical order and the correspondence between sensitive scenarios and time tiers, the historical time tier corresponding to each historical order is determined.

[0067] Step 330: Based on the historical time tiers corresponding to each historical order, the relationship between delivery efficiency and ETA, and the relationship between order conversion rate and ETA, determine the time tier corresponding to each sensitive scenario when the objective function of the change in delivery efficiency, the change in order conversion rate, and the time tier is minimized, and obtain the correspondence between sensitive scenarios and time tiers.

[0068] The relationship between delivery efficiency and ETA can be learned through machine learning models based on historical data; similarly, the relationship between order conversion rate and ETA can also be learned through machine learning models based on historical data. The relationships between delivery efficiency and ETA, and between order conversion rate and ETA, can be modeled using probability distributions, and their respective patterns of change with ETA can be derived based on learning from historical data. The change in delivery efficiency corresponds to the change in delivery efficiency when ETA is adjusted by a specific time interval, and the change in order conversion rate corresponds to the change in order conversion rate when ETA is adjusted by a specific time interval.

[0069] Figure 4a This is a schematic diagram illustrating the change in delivery efficiency with ETA in an embodiment of this application. Figure 4b This is a schematic diagram illustrating the change in order conversion rate with ETA in an embodiment of this application, as shown below. Figure 4a and Figure 4b (Simply observe the trend of the curve in the graph) As shown, ETA, as a crucial parameter for delivery, affects both delivery efficiency and order conversion rate. With increasing ETA, delivery efficiency significantly increases, while order conversion rate significantly decreases. Therefore, by adjusting ETA, an optimal time frame can be found to maximize delivery efficiency while minimizing the impact on order conversion rate.

[0070] The objective function is expressed as follows:

[0071] regret=∑(w1*ΔC+w2*ΔD+w3*ΔETA)

[0072] Where regret represents the value of the objective function, ΔC is the change in order conversion rate, w1 is the weight corresponding to the change in order conversion rate, ΔD is the change in delivery efficiency, w2 is the weight corresponding to the change in delivery efficiency, ΔETA is the time interval, and w3 is the weight corresponding to the time interval of ETA.

[0073] In sensitive scenarios corresponding to historical orders, Operations Research (OR) can be used to find the optimal time slot for each sensitive scenario. This involves selecting a time slot different from the historical time slots to minimize the global objective function value, which in turn minimizes the objective function value for all historical orders in the historical data, thus obtaining the optimal solution and the optimal time slot for each sensitive scenario. Heuristic algorithms (such as genetic algorithms and ALNS algorithms) can be used to modify the time slot for each historical order's sensitive scenario. For each historical order, the change in delivery efficiency as a function of ETA is obtained, along with the change in order conversion rate as a function of ETA. Based on the time slot, the change in delivery efficiency, and the change in order conversion rate, the objective function value for a historical order is obtained. The global objective function value for all historical orders in the historical data is then calculated by summing the objective function values ​​of all historical orders. Finally, the time slot corresponding to each sensitive scenario that minimizes the global objective function value is determined, establishing the correspondence between sensitive scenarios and time slots. The correspondence between sensitive scenarios and time slots determined by regret loss is the optimal correspondence. Applying this correspondence to online adjustments to ETA can significantly improve delivery efficiency while keeping order volume constant.

[0074] When optimizing the correspondence between sensitive scenarios and time slots, the thresholds of the delivery efficiency impact factor and the order conversion rate impact factor corresponding to each sensitive scenario can also be optimized simultaneously. This is achieved by using heuristic algorithms (such as genetic algorithms or ALNS algorithms) to simultaneously modify the time slot, delivery efficiency impact factor threshold, and order conversion rate impact factor threshold for each historical order's sensitive scenario. The goal is to minimize these thresholds in the objective function, thereby determining the time slot, delivery efficiency impact factor threshold, and order conversion rate impact factor threshold for each sensitive scenario. When adjusting the ETA online, these thresholds are used to determine the current sensitive scenario for the current order, and the corresponding time slot is determined based on the correspondence between sensitive scenarios and time slots. This achieves simultaneous optimization of the time slot and the thresholds of each impact factor.

[0075] Step 340: Obtain the ETA and order-related information for the current order.

[0076] Step 350: Based on the order-related information, determine the current sensitive scenario corresponding to the current order. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate.

[0077] Step 360: Determine the current time frame to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time frame.

[0078] Step 370: Adjust the ETA according to the current time range and output the adjusted ETA.

[0079] The estimated delivery time adjustment method provided in this embodiment determines the sensitive scenario corresponding to each historical order based on the order-related information of each historical order in historical data. It then determines the historical time tier corresponding to each historical order based on the sensitive scenario. Finally, based on the historical time tier, the relationship between delivery efficiency and ETA, and the relationship between order conversion rate and ETA, it determines the time tier corresponding to each sensitive scenario when the objective function of the changes in delivery efficiency, order conversion rate, and time tier is minimized. This yields the correspondence between sensitive scenarios and time tiers. Since the time tier corresponding to each sensitive scenario when the objective function is minimized is used as the correspondence between sensitive scenarios and time tiers, the globally optimal correspondence between sensitive scenarios and time tiers is obtained. Therefore, applying this method to online ETA adjustments can significantly improve delivery efficiency while maintaining a similar order size.

[0080] Based on the above technical solution, the method may further include: for orders of a first preset proportion, after determining the current time slot of the ETA, using the current time slot as the actual time slot, and adjusting the ETA according to the actual time slot; for orders of a second preset proportion, after determining the current time slot of the ETA, determining a time slot other than the current time slot as the actual time slot of the ETA, and adjusting the ETA according to the actual time slot; optimizing the correspondence between sensitive scenarios and time slots based on the sensitive scenarios and actual time slots corresponding to the orders of the first preset proportion, and the sensitive scenarios and actual time slots corresponding to the orders of the second preset proportion; wherein, the first preset proportion is greater than the second preset proportion, and the sum of the first preset proportion and the second preset proportion is equal to 1.

[0081] Due to the existence of operations research optimization, in certain sensitive scenarios, not all time slots have the opportunity to be selected. This can be addressed through the EE (Exploration and Exploitation) mechanism, which allocates a portion of the traffic for appropriate exploration, while the remaining traffic continues to be used for exploration. This effectively expands the upper limit of the optimization strategy. Since the feedback samples of ETA are not as expensive as those for pricing, EE exploration is sufficiently feasible. In this embodiment, exploration is based on the known best strategy, developing and utilizing time slots with known high returns; exploration does not consider past experience, probing for potentially high-return time slots.

[0082] When determining the time slot corresponding to the ETA of an order, the first preset proportion of orders are assigned their current time slots for ETA in the manner described above. These current time slots are then used as the actual time slots, and the ETAs are adjusted accordingly to ensure that the first preset proportion of orders use the optimal time slots corresponding to existing sensitive scenarios. For the second preset proportion of orders, after determining their current time slots for ETA in the manner described above, a time slot is randomly selected from those other than the current time slot as the actual time slot. The ETAs of these actual time slots are then adjusted to explore whether there are better time slots corresponding to sensitive scenarios. During operations optimization, based on the sensitive scenarios and actual time slots corresponding to the first preset proportion of orders, the sensitive scenarios and actual time slots corresponding to the second preset proportion of orders, the relationship between delivery efficiency and ETA, and the relationship between order conversion rate and ETA, the time slots corresponding to each order are modified. The time slot corresponding to each sensitive scenario is determined when the objective function for the changes in delivery efficiency, order conversion rate, and time slot is minimized, thus obtaining the correspondence between sensitive scenarios and time slots. By adjusting the ETA (Earnings To Delivery) for a second preset proportion of orders using a time slot other than the current one, optimal time slots for sensitive scenarios can be identified in a timely manner. The relationships between delivery efficiency and ETA, and between order conversion rate and ETA, are derived from historical data using machine learning models. These relationships are represented as stress curves, respectively. Furthermore, the estimated stress curve can be corrected based on the actual stress curve changes obtained using the EE (Extended Estimation) mechanism.

[0083] The first preset ratio is much larger than the second preset ratio. For example, the first preset ratio is 95% and the second preset ratio is 5%. This allows for the exploration of potentially better time slots for sensitive scenarios while ensuring the existing optimal correspondence between sensitive scenarios and time slots.

[0084] Example 3

[0085] This embodiment provides an estimated delivery time adjustment device, such as... Figure 5 As shown, the estimated delivery time adjustment device 500 includes:

[0086] The order information acquisition module 510 is configured to acquire the estimated delivery time (ETA) and order-related information for the current order.

[0087] The scenario determination module 520 is configured to determine the current sensitive scenario corresponding to the current order based on the ETA and the order-related information. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate.

[0088] The time tier determination module 530 is configured to determine the current time tier to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time tier, wherein the correspondence between the sensitive scenario and the time tier is determined in advance based on historical data.

[0089] ETA adjustment module 540 is configured to adjust the ETA according to the current time range and output the adjusted ETA.

[0090] Optionally, the device further includes:

[0091] The historical scenario determination module is configured to determine the sensitive scenario corresponding to each historical order based on the order-related information of each historical order in the historical data;

[0092] The historical time tier determination module is configured to determine the historical time tier corresponding to each historical order based on the sensitive scenario corresponding to each historical order.

[0093] The correspondence determination module is configured to determine the time slot corresponding to each sensitive scenario when the objective function of the change in delivery efficiency, the change in order conversion rate, and the time slot is minimized based on the historical time slot corresponding to each historical order, the relationship between delivery efficiency and ETA, and the relationship between order conversion rate and ETA, thereby obtaining the correspondence between sensitive scenarios and time slots.

[0094] Optionally, the objective function is expressed as follows:

[0095] regret=∑(w1*ΔC+w2*ΔD+w3*ΔETA)

[0096] Where regret represents the value of the objective function, ΔC is the change in order conversion rate, w1 is the weight corresponding to the change in order conversion rate, ΔD is the change in delivery efficiency, w2 is the weight corresponding to the change in delivery efficiency, ΔETA is the time interval, and w3 is the weight corresponding to the time interval of ETA.

[0097] Optionally, the scene determination module includes:

[0098] The influencing factor determination unit is configured to determine the delivery efficiency influencing factor value and the order conversion rate influencing factor value corresponding to the ETA based on the ETA and the order-related information.

[0099] The scenario determination unit is configured to determine the current sensitivity scenario corresponding to the current order based on the delivery efficiency impact factor value and the order conversion rate impact factor value.

[0100] Optionally, the scene determination unit is specifically used for:

[0101] Based on the relationship between the delivery efficiency impact factor value and the corresponding threshold, and the relationship between the order conversion rate impact factor value and the corresponding threshold, the current sensitivity scenario corresponding to the current order is determined.

[0102] Optionally, the device further includes:

[0103] The uncontrollable factor determination module is configured to determine the uncontrollable factors corresponding to the current order;

[0104] The scene determination module is specifically used for:

[0105] Based on the ETA, the order-related information, and uncontrollable factors, determine the current sensitivity scenario corresponding to the current order.

[0106] Optionally, the device further includes:

[0107] The first ETA adjustment module is configured to, after determining the current time range of the ETA for orders of the first preset ratio, use the current time range as the actual time range and adjust the ETA according to the actual time range.

[0108] The second ETA adjustment module is configured to determine, after determining the current time range of the ETA for orders of the second preset ratio, a time range other than the current time range as the actual time range of the ETA, and adjust the ETA according to the actual time range.

[0109] The correspondence optimization module is configured to optimize the correspondence between sensitive scenarios and time slots based on the sensitive scenarios and actual time slots corresponding to the first preset proportion of orders, and the sensitive scenarios and actual time slots corresponding to the second preset proportion of orders.

[0110] Wherein, the first preset ratio is greater than the second preset ratio, and the sum of the first preset ratio and the second preset ratio is equal to 1.

[0111] Optionally, the current order includes orders currently to be generated or orders currently generated.

[0112] Optionally, the ETA adjustment module includes:

[0113] The ETA output unit is configured to output the adjusted ETA to a client so that the adjusted ETA can be displayed on the client's order submission page or order details page.

[0114] The estimated delivery time adjustment device provided in this application embodiment is used to implement the steps of the estimated delivery time adjustment method described in this application embodiment. The specific implementation of each module of the device is described in the corresponding steps, and will not be repeated here.

[0115] The estimated delivery time adjustment device provided in this application embodiment obtains the ETA and order-related information of the current order, determines the current sensitive scenario corresponding to the current order based on the ETA and order-related information, determines the current time range of the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time range, adjusts the ETA based on the current time range, and outputs the adjusted ETA. Since the current sensitive scenario is related to delivery efficiency and order conversion rate, the current sensitive scenario obtained is a result that comprehensively considers delivery efficiency and order conversion rate, thereby improving delivery efficiency while ensuring order volume.

[0116] Example 4

[0117] This application also provides an electronic device, such as... Figure 6 As shown, the electronic device 600 may include one or more processors 610 and one or more memories 620 connected to the processors 610. The electronic device 600 may also include an input interface 630 and an output interface 640 for communicating with another device or system. Program code executed by the processor 610 may be stored in the memory 620.

[0118] The processor 610 in the electronic device 600 calls the program code stored in the memory 620 to execute the estimated delivery time adjustment method in the above embodiment.

[0119] The aforementioned components in the electronic device can be connected to each other via a bus, such as a data bus, address bus, control bus, expansion bus, and local bus, or any combination thereof.

[0120] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the estimated delivery time adjustment method as described in Embodiment 1 of this application.

[0121] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus embodiments, since they are fundamentally similar to the method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0122] The foregoing has provided a detailed description of a method, apparatus, electronic device, and storage medium for adjusting the estimated delivery time provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

[0123] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

Claims

1. A method for adjusting the estimated delivery time, characterized in that, include: Get the estimated delivery time (ETA) and order-related information for the current order; Based on the ETA and the order-related information, determine the current sensitive scenario corresponding to the current order. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate. Based on the current sensitive scenario and the correspondence between the sensitive scenario and the time tier, the current time tier to be adjusted for the ETA is determined, wherein the correspondence between the sensitive scenario and the time tier is determined in advance based on historical data. Adjust the ETA according to the current time range and output the adjusted ETA; Specifically, for orders with the first preset ratio, after determining the current time bracket of the ETA, the current time bracket is used as the actual time bracket, and the ETA is adjusted according to the actual time bracket. For orders with a second preset ratio, after determining the current time slot of the ETA, a time slot other than the current time slot is determined as the actual time slot of the ETA, and the ETA is adjusted according to the actual time slot. Based on the sensitive scenarios and actual time slots corresponding to the first preset proportion of orders, and the sensitive scenarios and actual time slots corresponding to the second preset proportion of orders, the correspondence between sensitive scenarios and time slots is optimized. Wherein, the first preset ratio is greater than the second preset ratio, and the sum of the first preset ratio and the second preset ratio is equal to 1.

2. The method according to claim 1, characterized in that, The correspondence between the sensitive scenarios and time slots is determined in advance based on historical data, including: Based on the order-related information of each historical order in the historical data, determine the sensitive scenarios corresponding to each historical order; Based on the sensitive scenarios corresponding to each historical order, determine the corresponding historical time frame for each historical order; Based on the historical time tiers corresponding to each historical order, the relationship between delivery efficiency and ETA, and the relationship between order conversion rate and ETA, we determine the time tier corresponding to each sensitive scenario when the objective function of the change in delivery efficiency, the change in order conversion rate, and the time tier is minimized, thus obtaining the correspondence between sensitive scenarios and time tiers.

3. The method according to claim 2, characterized in that, The objective function is expressed as follows: regret=∑(w_1* C+w_2* D+w_3* ETA) Where regret represents the value of the objective function. C represents the change in order conversion rate, and w_1 represents the weight corresponding to the change in order conversion rate. D represents the change in delivery efficiency, and w_2 represents the weight corresponding to the change in delivery efficiency. ETA represents the time tier, and w_3 represents the weight corresponding to the time tier of ETA.

4. The method according to any one of claims 1-3, characterized in that, The step of determining the current sensitivity scenario corresponding to the current order based on the ETA and the order-related information includes: Based on the ETA and the order-related information, determine the delivery efficiency impact factor value and order conversion rate impact factor value corresponding to the ETA; Based on the delivery efficiency impact factor value and the order conversion rate impact factor value, determine the current sensitivity scenario corresponding to the current order.

5. The method according to claim 4, characterized in that, The step of determining the current sensitivity scenario corresponding to the current order based on the delivery efficiency impact factor value and the order conversion rate impact factor value includes: Based on the relationship between the delivery efficiency impact factor value and the corresponding threshold, and the relationship between the order conversion rate impact factor value and the corresponding threshold, the current sensitivity scenario corresponding to the current order is determined.

6. The method according to any one of claims 1-3, characterized in that, Also includes: Identify the uncontrollable factors corresponding to the current order; The step of determining the current sensitivity scenario corresponding to the current order based on the ETA and the order-related information includes: Based on the ETA, the order-related information, and uncontrollable factors, determine the current sensitivity scenario corresponding to the current order.

7. The method according to any one of claims 1-3, characterized in that, The current order includes orders that are currently pending or orders that have already been generated.

8. The method according to any one of claims 1-3, characterized in that, The adjusted ETA output includes: The adjusted ETA is output to the client so that it can be displayed on the client's order submission page or order details page.

9. A device for adjusting the estimated delivery time, characterized in that, include: The order information acquisition module is configured to acquire the estimated delivery time (ETA) and related order information for the current order. The scenario determination module is configured to determine the current sensitive scenario corresponding to the current order based on the ETA and the order-related information. The current sensitive scenario is a scenario related to delivery efficiency and order conversion rate. The time tier determination module is configured to determine the current time tier to be adjusted for the ETA based on the current sensitive scenario and the correspondence between the sensitive scenario and the time tier, wherein the correspondence between the sensitive scenario and the time tier is determined in advance based on historical data; wherein, for orders of a first preset proportion, after determining the current time tier of the ETA, the current time tier is used as the actual time tier, and the ETA is adjusted according to the actual time tier; for orders of a second preset proportion, after determining the current time tier of the ETA, a time tier other than the current time tier is determined as the actual time tier of the ETA, and the ETA is adjusted according to the actual time tier; the correspondence between sensitive scenarios and time tiers is optimized based on the sensitive scenarios and actual time tiers corresponding to the orders of the first preset proportion and the orders of the second preset proportion; wherein, the first preset proportion is greater than the second preset proportion, and the sum of the first preset proportion and the second preset proportion is equal to 1; The ETA adjustment module is configured to adjust the ETA according to the current time range and output the adjusted ETA.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for adjusting the estimated delivery time as described in any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method for adjusting the estimated delivery time as described in any one of claims 1 to 8.

Images