Method, apparatus and device for personalizing configuration of a signing mode and medium

Abstract

This invention provides a method, apparatus, device, and medium for personalized configuration of delivery confirmation methods, including: constructing a multi-dimensional attribute library for packages; extracting package attributes corresponding to a target package based on the constructed multi-dimensional attribute library; constructing a feature vector for the target package based on the package attributes of the target package and the logical relationships between the attributes; and completing intelligent matching and decision-making of delivery confirmation rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and a digital delivery confirmation rule library. This invention addresses industry pain points such as rigid processes, insufficient intelligence, significant system performance shortcomings, and poor adaptability to complex scenarios in traditional logistics delivery confirmation models. Through a complete technical system encompassing multi-dimensional attribute modeling, weighted intelligent matching, LLM for complex scenarios, high-frequency rule caching, and high-concurrency traffic control, it achieves precise personalized configuration of delivery confirmation methods and efficient and stable system operation.

Description

Technical Field

[0001] This invention relates to the field of logistics management technology, and more specifically, to methods, apparatus, equipment, and media for personalized configuration of receipt confirmation methods. Background Technology

[0002] With the rapid development of the e-commerce and express delivery industries, parcel delivery volume continues to climb. Traditional express delivery signing models mostly adopt a unified signing process, failing to differentiate between different types of parcels, different user groups, and different delivery scenarios, resulting in numerous industry pain points. On the one hand, high-value parcels, fresh and fragile items, and other special categories of parcels are prone to problems such as lost packages, difficulty in tracing damage, and non-standard acceptance when using ordinary signing methods, while strict signing processes for ordinary daily necessities would reduce delivery efficiency. On the other hand, the signing needs of VIP users and ordinary users, urban and remote areas, and regular and high-concurrency periods during peak sales periods differ significantly, and fixed signing models cannot simultaneously address delivery safety, efficiency, and user experience. Furthermore, existing signing rule matching methods are mostly fixed, hard-coded rules, lacking flexibility. Repeated matching calculations of high-frequency rules consume system resources, easily leading to system lag and response delays during high-concurrency periods. Moreover, existing rule bases cannot cover complex scenarios with cross-attribute combinations, making it difficult to generate suitable signing solutions. This results in low intelligence and poor personalization in the logistics signing process, failing to meet the current development needs of the logistics industry for refined and intelligent delivery. Summary of the Invention

[0003] The main objective of this invention is to solve the problems of poor flexibility in existing signing rules matching methods, low level of intelligence in the logistics signing process, and poor personalization adaptability.

[0004] The first aspect of this invention provides a method for personalized configuration of receipt confirmation methods, comprising: Build a multi-dimensional attribute library for packages; Based on the constructed multi-dimensional attribute library of packages, the package attributes corresponding to the target package are extracted. Based on the package attributes of the target package and the logical relationship between each attribute, the feature vector of the target package is constructed. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed.

[0005] Optionally, in a first implementation of the first aspect of the present invention, constructing the multi-dimensional attribute library of the package includes: The system sets up multi-dimensional attributes including e-commerce platform, product type, regional code, delivery time and user level; it configures an adjustable value range or attribute type for each dimension attribute; and it configures a corresponding weight for each dimension attribute, which is then integrated to form a global attribute weight.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the step of intelligently matching and deciding on the signing rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, includes: Establish a digital signature rule library; the digital signature rule library is a set of identifiable standardized signature rules built on a multi-dimensional attribute library of packages; Based on global attribute weights, the weighted similarity between the feature vector of the target package and each signing rule in the digital signing rule base is calculated. The acceptance rule with the highest weighted similarity score and the highest similarity score is selected as the final acceptance rule for the target package. If there are multiple acceptance rules with a weighted similarity greater than the threshold and the same weighted similarity score, the optimal acceptance rule will be selected through a preset rule priority system. If the weighted similarity scores are all less than the threshold, it is considered that the current target package has not matched the corresponding signing rule, and the LLM inference model is triggered to generate personalized signing rules.

[0007] Optionally, in a third implementation of the first aspect of the present invention, triggering the LLM inference model to generate personalized acceptance rules includes: the LLM inference model generating corresponding personalized acceptance rules based on the feature information of the target package, a preset business knowledge base, and preset inference logic. The feature information of the target package includes: the feature vector of the target package, the global attribute weight, and the scene anomaly features of the currently unmatched rule; The preset business knowledge base includes: high-value package signing specifications, fresh or fragile item signing requirements, VIP user service standards, security verification rules, and regional delivery policies. The preset reasoning logic includes: security priority principle, efficiency priority principle, weight priority principle, multiple verification rules, and fallback rules for abnormal scenarios.

[0008] Optionally, in the fourth implementation of the first aspect of the present invention, the method further includes: locally caching the signature rules whose matching retrieval frequency is greater than a preset frequency threshold; for a signature rule matching request, prioritizing matching the cached signature rules from the local cache; and when a match is found, retrieving the corresponding matching result from the local cache; wherein the local cache adopts a cache update mechanism of timed update and / or triggered update.

[0009] Optionally, in the fifth implementation of the first aspect of the present invention, the method further includes: acquiring and statistically analyzing the traffic data of the acceptance requests in real time; triggering a flow limiting mechanism when the traffic of the acceptance requests exceeds a preset traffic threshold; queuing the acceptance requests that exceed the capacity in an orderly manner; grouping the batch acceptance requests; processing the rule matching or LLM inference operation of each group of requests using an asynchronous calculation method; and returning the corresponding acceptance method decision result according to the original request order after the calculation is completed.

[0010] Optionally, in the sixth implementation of the first aspect of the present invention, the method further includes: abstracting and encapsulating various signing methods, including taking pictures, positioning, and signing code verification, into independent capability components; wherein each capability component is encapsulated with exclusive functional logic, input and output parameters and execution dependency conditions, and configured with a unique component identifier and registration information; When the client app starts, each capability component registers with the central router built into the client. The central router receives and records the identifier, function, availability status, and dependencies of each component, generating a component registration list. After receiving the signing method decision result from the server, the client parses the target signing method combination and extracts the corresponding component identifier, then sends a component call request to the central router. The central router verifies the availability and dependencies of the components based on the component registration list. If the verification is successful, the corresponding component is dynamically instantiated. Then, according to the preset execution order of the decision result, the instantiated components are orchestrated, defining the execution order, data transmission rules, and exception handling logic, generating an intelligent signing task pipeline, driving each component to complete the corresponding signing operation according to the pipeline.

[0011] A second aspect of the present invention provides an apparatus for personalized configuration of receipt methods, comprising: The package attribute library building module is used to build a multi-dimensional attribute library for packages; The package feature vector construction module is used to extract the package attributes corresponding to the target package based on the constructed multi-dimensional package attribute library, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between the attributes. The signature rule matching module is used to intelligently match and decide on signature rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signature rule library.

[0012] Optionally, in a first implementation of the second aspect of the present invention, constructing the multi-dimensional attribute library of the package includes: The system sets up multi-dimensional attributes including e-commerce platform, product type, regional code, delivery time and user level; it configures an adjustable value range or attribute type for each dimension attribute; and it configures a corresponding weight for each dimension attribute, which is then integrated to form a global attribute weight.

[0013] Optionally, in a second implementation of the second aspect of the present invention, the signature rule matching module includes: The receipt rule base construction submodule is used to build a digital receipt rule base; the digital receipt rule base is a set of identifiable standardized receipt rules built on a multi-dimensional attribute library of packages. The weighted similarity calculation submodule is used to calculate the weighted similarity between the feature vector of the target package and each signing rule in the digital signing rule base, based on the global attribute weights. The rule matching submodule is used to select the signing rule with the highest weighted similarity score and a weighted similarity score greater than the threshold as the final signing rule for the target package. If there are multiple signing rules with the highest weighted similarity scores and weighted similarity scores, the optimal signing rule is selected through a preset rule priority system. If the weighted similarity scores are all less than the threshold, it is considered that the current target package has not matched the corresponding signing rule, and the LLM inference model is triggered to generate a personalized signing rule.

[0014] Optionally, in a third implementation of the second aspect of the present invention, the step of triggering the LLM inference model to generate personalized acceptance rules includes: the LLM inference model generating corresponding personalized acceptance rules based on the feature information of the target package, a preset business knowledge base, and preset inference logic. The feature information of the target package includes: the feature vector of the target package, the global attribute weight, and the scene anomaly features of the currently unmatched rule; The preset business knowledge base includes: high-value package signing specifications, fresh or fragile item signing requirements, VIP user service standards, security verification rules, and regional delivery policies. The preset reasoning logic includes: security priority principle, efficiency priority principle, weight priority principle, multiple verification rules, and fallback rules for abnormal scenarios.

[0015] Optionally, in the fourth implementation of the second aspect of the present invention, the system further includes: a cache optimization module, used to locally cache the signature rules whose signature rule matching retrieval frequency is greater than a preset frequency threshold, and for a signature rule matching request, prioritize matching the cached signature rules from the local cache, and when a match is found, retrieve the corresponding matching result from the local cache; wherein, the local cache adopts a cache update mechanism of timed update and / or triggered update.

[0016] Optionally, in the fifth implementation of the second aspect of the present invention, the system further includes: a high-concurrency optimization module, used to acquire and count the traffic data of the acceptance requests in real time; when the traffic of the acceptance requests exceeds a preset traffic threshold, a rate limiting mechanism is triggered to queue the acceptance requests that exceed the capacity in an orderly manner; at the same time, the batch acceptance requests are grouped and the rule matching or LLM inference operation of each group of requests is processed in an asynchronous calculation manner; after the calculation is completed, the corresponding acceptance method decision result is returned in the original request order.

[0017] Optionally, in the sixth implementation of the second aspect of the present invention, the system further includes: a signature execution component module, used to abstract and encapsulate various signature methods, including taking pictures, positioning, and signature code verification, into independent capability components; wherein, each capability component is encapsulated with exclusive functional logic, input and output parameters and execution dependency conditions, and configured with a unique component identifier and registration information; When the client app starts, each capability component registers with the central router built into the client. The central router receives and records the identifier, function, availability status, and dependencies of each component, generating a component registration list. After receiving the signing method decision result from the server, the client parses the target signing method combination and extracts the corresponding component identifier, then sends a component call request to the central router. The central router verifies the availability and dependencies of the components based on the component registration list. If the verification is successful, the corresponding component is dynamically instantiated. Then, according to the preset execution order of the decision result, the instantiated components are orchestrated, defining the execution order, data transmission rules, and exception handling logic, generating an intelligent signing task pipeline, driving each component to complete the corresponding signing operation according to the pipeline.

[0018] A third aspect of the present invention provides an electronic device, the electronic device comprising a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the electronic device to perform the steps of the method for personalized configuration of the receipt method as described above.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the method for personalized configuration of receipt methods as described above.

[0020] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention achieves precise and differentiated configuration of signing methods in different scenarios by using multi-dimensional attribute modeling of packages and global weighted matching, thus balancing signing security and delivery efficiency. 2. This invention achieves comprehensive decision-making in complex and special scenarios by using weighted similarity threshold judgment and LLM large model fallback reasoning, thus overcoming the coverage limitations of fixed rule bases. 3. This invention achieves rapid response to regular requests and reduces the waste of redundant computing power by using the technical features of local caching of high-frequency signing rules and dual cache updates; 4. This invention achieves stable system operation and efficient request processing in high-concurrency scenarios through the technical features of real-time traffic monitoring, threshold-triggered rate limiting, orderly request queuing, asynchronous grouping calculation, and result reordering. 5. This invention addresses industry pain points such as rigid processes, insufficient intelligence, prominent system performance shortcomings, and poor adaptability to complex scenarios in traditional logistics signing and receiving models. Through a complete technical system including multi-dimensional attribute modeling, weighted intelligent matching, LLM complex scenario fallback, high-frequency rule caching, and high-concurrency traffic control, it achieves precise and personalized configuration of signing and receiving methods and efficient and stable system operation. Attached Figure Description

[0021] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 The first flowchart of a method for personalized configuration of receipt methods provided in an embodiment of the present invention.

[0022] Figure 2 The second flowchart is provided for a method of personalized configuration of receipt method in an embodiment of the present invention.

[0023] Figure 3 The third flowchart of the method for personalized configuration of receipt method provided in the embodiments of the present invention.

[0024] Figure 4 The fourth flowchart of the method for personalized configuration of receipt method provided in the embodiments of the present invention.

[0025] Figure 5 This is a schematic diagram of a device for configuring a personalized receipt method according to an embodiment of the present invention.

[0026] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0027] This invention provides a method, apparatus, device, and medium for personalized configuration of delivery receipt methods, including: constructing a multi-dimensional attribute library for packages; extracting package attributes corresponding to a target package based on the constructed multi-dimensional attribute library; constructing a feature vector for the target package based on the package attributes of the target package and the logical relationships between the attributes; and completing intelligent matching and decision-making of delivery receipt rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and a digital delivery receipt rule library. This invention solves the problems of poor flexibility, low intelligence level, and poor personalization adaptability in existing delivery receipt rule matching methods.

[0028] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 The first embodiment of the method for personalized configuration of receipt methods in this invention includes: 101. Construct a multi-dimensional attribute library for packages; In this embodiment, e-commerce platform, product type, regional code, delivery time and user level are set as multi-dimensional attributes; an adjustable value range or attribute type is configured for each dimension attribute; at the same time, a corresponding weight is configured for each dimension attribute, and they are integrated to form a global attribute weight.

[0030] The e-commerce platforms include: self-operated platforms, third-party merchant platforms, and cross-border e-commerce platforms; the product types include: general daily necessities, high-value 3C products, fresh and perishable goods, fragile goods, and prohibited and controlled items; the area codes include: urban areas, suburbs, remote towns, specially controlled areas, and residential communities / stations / self-pickup points; the delivery times include: regular delivery, express delivery, same-day / next-day delivery, and time-limited delivery for fresh produce; and the user levels include: ordinary users, VIP members, and corporate clients. For each of the above-mentioned dimension attributes, an adjustable value range or standardized attribute type can be configured separately. For example, the product type attribute can be set with a clear category enumeration value, the regional code attribute can be connected to the national standard regional code library, and the user level attribute can be set with a hierarchical division standard. All values ​​can be flexibly adjusted according to business expansion to ensure the scalability of the attribute library.

[0031] The step of configuring corresponding weights for each dimension attribute and integrating them to form global attribute weights includes: configuring a corresponding matching weight for each dimension attribute individually, with the weight value reflecting the degree of influence of the attribute on the acceptance decision; the more critical the attribute, the higher the corresponding weight value; and integrating and normalizing the independent weights of each dimension attribute to form global attribute weights. This weight is a globally universal configuration that can be adjusted and optimized periodically according to business rules and used for subsequent weighted similarity calculations to highlight the decision-making influence of core attributes and avoid secondary attributes interfering with the decision results.

[0032] 102. Based on the constructed multi-dimensional attribute library of packages, extract the package attributes corresponding to the target package, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between each attribute; This embodiment transforms unstructured package features into standardized, computable digital feature vectors, realizing the quantitative expression of package features and providing a data foundation for subsequent similarity matching calculations.

[0033] 103. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed.

[0034] In this embodiment, a digital signature rule library is built based on the attribute system of the multi-dimensional attribute database of packages. This rule library is a standardized set of signature rules that the system can directly identify and execute. Each rule corresponds to a signature method for a certain type of package scenario. The rule content includes standardized fields such as attribute matching conditions, signature verification methods, and execution requirements, which correspond one-to-one with package attributes to avoid rule ambiguity. The rule library can be pre-entered, imported in batches, or dynamically updated later to adapt to business rule iterations.

[0035] The global attribute weights are invoked to perform item-by-item similarity calculations between the feature vector of the target package and each stored signature rule in the digital signature rule library. During the calculation process, the weight values ​​of the corresponding attributes are introduced to carry out weighted similarity calculations, rather than simple equal similarity calculations. The core purpose is to amplify the impact of high-weight attributes on the matching results and weaken the interference of low-weight attributes, so as to ensure that the matching results are in line with business priorities.

[0036] After calculating the weighted similarity for all rules, the final selection is completed according to a three-layer logic: First layer: Threshold filtering, only retaining the signing rules with a weighted similarity greater than the preset threshold, and filtering out invalid rules with low matching degree or that are not applicable; The second layer: optimal screening. Among the rules above the threshold, the rule with the highest weighted similarity score is selected and directly used as the final acceptance rule for the target package. The third layer: tie-breaking. If multiple rules meet the condition of being greater than the threshold and having the highest score, the rules are not randomly selected. Instead, a preset rule priority system is used for further filtering. The priority is set according to business specifications, and finally, a unique and optimal acceptance rule is determined.

[0037] If the weighted similarity score of all signing rules is lower than the preset threshold, it indicates that the target package is a complex case and the existing standardized rule base cannot cover the matching. The system automatically determines that the target package does not match the corresponding signing rule. At this time, the LLM inference model is directly triggered to generate personalized signing rules adapted to the complex scenario based on the LLM inference model.

[0038] The personalized acceptance rules generated based on the LLM inference model to adapt to this complex scenario include: After triggering the LLM inference model, three types of core feature information are input into the LLM inference model, including: the feature vector of the target package, the global attribute weights, and the scene anomaly features of the currently unmatched rules.

[0039] The LLM inference model performs inference based on two pre-set criteria to ensure that the generated rules comply with logistics industry standards and business requirements, and that there are no illegal or unreasonable decisions: First, it has a pre-set business knowledge base with built-in logistics signing standards for all scenarios, including high-value parcel signing standards, fresh or fragile item signing requirements, VIP user service standards, security verification rules, and regional delivery policies. Second, it pre-sets reasoning logic and incorporates standardized decision-making principles, including the principles of safety priority, efficiency priority, weight priority, multiple verification rules, and fallback rules for abnormal scenarios.

[0040] The LLM inference model combines the input package feature information with a pre-set business knowledge base and inference logic to perform semantic understanding, scenario analysis and logical reasoning. It automatically generates personalized receipt rules that are adapted to the current complex scenario. The rules include clear receipt verification methods, execution processes and security requirements. They can be directly sent to the client for execution and are completed automatically without the need for manual rule writing.

[0041] Please see Figure 2 A second embodiment of the method for personalized configuration of receipt methods in this invention includes: 201. Construct a multi-dimensional attribute library for packages; 202. Based on the constructed multi-dimensional attribute library of packages, extract the package attributes corresponding to the target package, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between each attribute; 203. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed. 204. Locally cache the signature rules that have a matching frequency greater than a preset frequency threshold. For signature rule matching requests, prioritize matching the cached signature rules in the local cache. When a match is found, retrieve the corresponding matching result from the local cache. In this embodiment, the matching and retrieval frequency of each receipt rule is counted in real time, and a dedicated rule retrieval frequency statistics ledger is established to dynamically track the number of rule calls in the digital receipt rule library. A standardized high-frequency retrieval frequency threshold is preset, which can be flexibly configured and adjusted according to the system's carrying capacity and the volume of business requests. When the matching and retrieval frequency of a certain receipt rule exceeds the preset high-frequency retrieval frequency threshold within a unit of time, the rule is automatically determined to be a high-frequency reuse rule, and the complete receipt decision result corresponding to the rule is synchronously stored in the local cache space to complete the cache storage operation.

[0042] For each initiated signature rule matching request, the execution order prioritizes cache and follows rule base matching, without requiring manual switching or intervention: First, the local cache space is automatically read, and the feature information of the current target package is quickly matched and verified with the high-frequency rules stored in the local cache; if a match is found, that is, if there is a signature rule in the local cache that is completely suitable for the current package, the corresponding signature rule decision result is immediately retrieved from the local cache, which greatly shortens the request response time and achieves millisecond-level fast decision-making.

[0043] To ensure that the acceptance rules in the cache are always consistent with the latest business specifications and system rules, and to avoid cached content from expiring, becoming invalid, or conflicting with existing rules, this embodiment adopts a dual cache update mechanism that combines scheduled updates and triggered updates, taking into account both cache timeliness and system stability.

[0044] Please see Figure 3 A third embodiment of the method for personalized configuration of receipt methods in this invention includes: 301. Construct a multi-dimensional attribute library for packages; 302. Based on the constructed multi-dimensional attribute library of packages, extract the package attributes corresponding to the target package, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between each attribute; 303. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed. 304. Real-time acquisition and statistics of the traffic data of the signing request. When the traffic of the signing request exceeds the preset traffic threshold, the rate limiting mechanism is triggered to queue the signing requests that exceed the capacity in an orderly manner. At the same time, the batch signing requests are grouped and the rule matching or LLM inference operation of each group of requests is processed in an asynchronous calculation method. After the calculation is completed, the corresponding signing method decision result is returned according to the original request order. In this embodiment, a dedicated real-time traffic monitoring module is deployed in the system backend to collect and statistically analyze all incoming requests continuously. Monitoring dimensions include core traffic metrics such as the total number of requests per unit time, concurrent requests, and request source ports. Statistical data is updated and dynamically refreshed in real time, ensuring no data delays or omissions. Traffic statistics run silently in the background without consuming frontend decision-making and execution resources. Simultaneously, traffic data is synchronized to the system control center in real time, providing accurate and real-time data support for subsequent rate limiting triggers, achieving second-level awareness of traffic fluctuations. Standardized traffic thresholds are pre-defined based on system hardware capacity, server computing power, and daily peak business volume. These thresholds represent the maximum number of signature requests the system can stably handle and can be flexibly adjusted according to system expansion and business volume growth to adapt to different operational needs. The real-time traffic monitoring module continuously compares the current statistical traffic with the preset thresholds. Once the current signature request traffic exceeds the preset threshold, the rate limiting mechanism is automatically triggered immediately. The entire process requires no manual operation and has no triggering delay, achieving second-level response control in high-concurrency scenarios. Once the rate limiting mechanism is triggered, requests exceeding the current capacity limit are not directly rejected, discarded, or disordered. Instead, a request queuing queue is initiated, and excess requests are placed into a standardized queue for orderly waiting. The queuing rules strictly follow the order of request arrival; requests arriving first are placed at the front of the queue, and subsequent requests are placed in order. Each request is uniquely identified and its original arrival time is marked to prevent queue jumping, out-of-order processing, and request loss. During the queuing process, the system releases idle computing power in real time and processes requests step by step according to the queue order, ensuring that the system load remains within a safe threshold and preventing overload and lag. While request queuing and rate limiting are in place, batch grouping processing logic is initiated for the batch acceptance requests that are currently pending. According to the preset grouping capacity, the large batch of requests is split into several small request groups. After grouping, asynchronous computation is used to process the request tasks within each group in parallel. Each group's tasks run independently without interfering with each other. The core processing tasks within each group include two types of core operations: regular rule matching calculation and LLM inference model operation. Asynchronous computation can make full use of the system's idle computing power, significantly improving overall processing efficiency and solving the problems of long processing times and low efficiency in synchronous computation. This embodiment addresses high-concurrency traffic scenarios such as e-commerce promotions, holidays, and concentrated delivery, resolving issues like system overload, lag, request loss, and response delays caused by a massive influx of delivery confirmation requests during peak periods. It balances system stability, request processing efficiency, and result orderliness without compromising decision accuracy. Please see Figure 4 The fourth embodiment of the method for personalized configuration of receipt methods in this invention includes: 401. Construct a multi-dimensional attribute library for packages; 402. Based on the constructed multi-dimensional attribute library of packages, extract the package attributes corresponding to the target package, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between each attribute; 403. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed. 404. Abstract the signing methods such as taking photos, positioning, and signing code verification into independent capability components. When the APP starts, the component registers with the central router. After the client receives the server's decision, it dynamically instantiates and orchestrates the component through the router to generate an intelligent signing task pipeline.

[0045] In this embodiment, various signing methods, including taking photos, positioning, and signing code verification, are abstracted and encapsulated into independent capability components. Each capability component includes dedicated functional logic, input and output parameters, and execution dependency conditions, and is configured with a unique component identifier and registration information. When the client app starts, each capability component registers with the central router built into the client. The central router receives and records the identifier, function, availability status, and dependencies of each component, generating a component registration list. After receiving the signing method decision result from the server, the client parses the target signing method combination and extracts the corresponding component identifier, then sends a component call request to the central router. The central router verifies the availability and dependencies of the components based on the component registration list. If the verification is successful, the corresponding component is dynamically instantiated. Then, according to the preset execution order of the decision result, the instantiated components are orchestrated, defining the execution order, data transmission rules, and exception handling logic, generating an intelligent signing task pipeline, driving each component to complete the corresponding signing operation according to the pipeline.

[0046] The method for personalized signature configuration in the embodiments of the present invention has been described above. The apparatus for personalized signature configuration in the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 5 One embodiment of the device for personalized configuration of receipt methods in this invention includes: Package Attribute Library Building Module 501 is used to build a multi-dimensional package attribute library; In this embodiment, the package attribute library construction module 501 includes: The system sets up multi-dimensional attributes including e-commerce platform, product type, regional code, delivery time and user level; it configures an adjustable value range or attribute type for each dimension attribute; and it configures a corresponding weight for each dimension attribute, which is then integrated to form a global attribute weight.

[0047] The package feature vector construction module 502 is used to extract the package attributes corresponding to the target package based on the constructed package multi-dimensional attribute library, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between the attributes. The signature rule matching module 503 is used to intelligently match and decide on signature rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signature rule library.

[0048] In this embodiment, the signature rule matching module 503 includes: The receipt rule base construction submodule 5031 is used to build a digital receipt rule base; the digital receipt rule base is a set of identifiable standardized receipt rules built based on a multi-dimensional attribute library of packages. The weighted similarity calculation submodule 5032 is used to calculate the weighted similarity between the feature vector of the target package and each signing rule in the digital signing rule base based on the global attribute weight. The rule matching submodule 5033 is used to select the signing rule with the highest weighted similarity score and a weighted similarity score greater than the threshold as the final signing rule for the target package. If there are multiple signing rules with the highest weighted similarity scores and weighted similarity scores, the optimal signing rule is selected through a preset rule priority system. If the weighted similarity scores are all less than the threshold, it is considered that the current target package has not matched the corresponding signing rule, and the LLM inference model is triggered to generate a personalized signing rule. The step of triggering the LLM inference model to generate personalized acceptance rules includes: the LLM inference model generates corresponding personalized acceptance rules based on the feature information of the target package, a preset business knowledge base, and preset inference logic. The feature information of the target package includes: the feature vector of the target package, the global attribute weight, and the scene anomaly features of the currently unmatched rule; The preset business knowledge base includes: high-value package signing specifications, fresh or fragile item signing requirements, VIP user service standards, security verification rules, and regional delivery policies. The preset reasoning logic includes: security priority principle, efficiency priority principle, weight priority principle, multiple verification rules, and fallback rules for abnormal scenarios.

[0049] The cache optimization module 504 is used to cache the receipt rules whose matching retrieval frequency is greater than a preset frequency threshold locally. For the receipt rule matching request, it first matches the cached receipt rules in the local cache. When a match is found, the corresponding matching result is retrieved from the local cache. The local cache adopts a cache update mechanism of timed update and / or triggered update. The high-concurrency optimization module 505 is used to acquire and count the traffic data of the signing requests in real time. When the traffic of the signing requests exceeds the preset traffic threshold, the rate limiting mechanism is triggered to queue the signing requests that exceed the capacity in an orderly manner. At the same time, the batch signing requests are grouped and the rule matching or LLM inference operation of each group of requests is processed in an asynchronous calculation method. After the calculation is completed, the corresponding signing method decision result is returned according to the original request order. The signature execution component module 505 is used to abstract and encapsulate various signature methods, including taking pictures, positioning, and signature code verification, into independent capability components. Each capability component includes exclusive functional logic, input and output parameters, and execution dependency conditions, and is configured with a unique component identifier and registration information. When the client app starts, each capability component registers with the central router built into the client. The central router receives and records the identifier, function, availability status, and dependencies of each component, generating a component registration list. After receiving the signing method decision result from the server, the client parses the target signing method combination and extracts the corresponding component identifier, then sends a component call request to the central router. The central router verifies the availability and dependencies of the components based on the component registration list. If the verification is successful, the corresponding component is dynamically instantiated. Then, according to the preset execution order of the decision result, the instantiated components are orchestrated, defining the execution order, data transmission rules, and exception handling logic, generating an intelligent signing task pipeline, driving each component to complete the corresponding signing operation according to the pipeline.

[0050] above Figure 5 The device for personalized configuration of receipt method in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The electronic device in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0051] Figure 6This is a schematic diagram of the structure of an electronic device 700 provided in an embodiment of the present invention. The electronic device 700 can vary significantly due to different configurations or performance characteristics. It may include one or more central processing units (CPUs) 710 (e.g., one or more processors) and a memory 720, and one or more storage media 730 (e.g., one or more mass storage devices) for storing application programs 733 or data 732. The memory 720 and storage media 730 can be temporary or persistent storage. The program stored in the storage media 730 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the electronic device 700. Furthermore, the processor 710 may be configured to communicate with the storage media 730 and execute the series of instruction operations in the storage media 730 on the electronic device 700.

[0052] Electronic device 700 may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input / output interfaces 750, and / or one or more operating systems 731, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 6 The illustrated electronic device structure does not constitute a limitation on electronic devices and may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0053] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of a method for personalized configuration acceptance.

[0054] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0055] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0056] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for personalized configuration of delivery confirmation methods, characterized in that, include: Build a multi-dimensional attribute library for packages; Based on the constructed multi-dimensional attribute library of packages, the package attributes corresponding to the target package are extracted. Based on the package attributes of the target package and the logical relationship between each attribute, the feature vector of the target package is constructed. Based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signing rule library, intelligent matching and decision-making of signing rules are completed.

2. The method for personalized configuration of receipt signing as described in claim 1, characterized in that, The construction of the multi-dimensional attribute library for packages includes: The system sets up multi-dimensional attributes including e-commerce platform, product type, regional code, delivery time and user level; it configures an adjustable value range or attribute type for each dimension attribute; and it configures a corresponding weight for each dimension attribute, which is then integrated to form a global attribute weight.

3. The method for personalized configuration of receipt signing as described in claim 1, characterized in that, The intelligent matching and decision-making of signing rules, based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute and the digital signing rule base, includes: Establish a digital signature rule library; the digital signature rule library is a set of identifiable standardized signature rules built on a multi-dimensional attribute library of packages; Based on global attribute weights, the weighted similarity between the feature vector of the target package and each signing rule in the digital signing rule base is calculated. The acceptance rule with the highest weighted similarity score and the highest similarity score is selected as the final acceptance rule for the target package. If there are multiple acceptance rules with a weighted similarity greater than the threshold and the same weighted similarity score, the optimal acceptance rule will be selected through a preset rule priority system. If the weighted similarity scores are all less than the threshold, it is considered that the current target package has not matched the corresponding signing rule, and the LLM inference model is triggered to generate personalized signing rules.

4. The method for personalized configuration of receipt signing method according to claim 3, characterized in that, The triggering of the LLM inference model to generate personalized acceptance rules includes: the LLM inference model generates corresponding personalized acceptance rules based on the feature information of the target package, a preset business knowledge base, and preset inference logic. The feature information of the target package includes: the feature vector of the target package, the global attribute weight, and the scene anomaly features of the currently unmatched rule; The preset business knowledge base includes: high-value package signing specifications, fresh or fragile item signing requirements, VIP user service standards, security verification rules, and regional delivery policies. The preset reasoning logic includes: security priority principle, efficiency priority principle, weight priority principle, multiple verification rules, and fallback rules for abnormal scenarios.

5. The method for personalized configuration of receipt signing method according to claim 3, characterized in that, The method further includes: caching local receipt rules whose matching frequency is greater than a preset frequency threshold; for receipt rule matching requests, prioritizing matching cached receipt rules from the local cache; and retrieving the corresponding matching result from the local cache when a match is found; wherein the local cache adopts a cache update mechanism of timed update and / or triggered update.

6. The method for personalized configuration of receipt signing method according to claim 3, characterized in that, The method further includes: acquiring and statistically analyzing the traffic data of the acceptance requests in real time; triggering a rate limiting mechanism when the traffic of the acceptance requests exceeds a preset traffic threshold; queuing the acceptance requests that exceed the capacity in an orderly manner; grouping the batch acceptance requests; processing the rule matching or LLM inference operation of each group of requests using an asynchronous calculation method; and returning the corresponding acceptance method decision result according to the original request order after the calculation is completed.

7. The method for personalized configuration of receipt signing method according to claim 1, characterized in that, The method further includes: abstracting and encapsulating various signing methods, including taking photos, positioning, and signing code verification, into independent capability components; wherein each capability component is encapsulated with exclusive functional logic, input and output parameters, and execution dependency conditions, and configured with a unique component identifier and registration information; When the client app starts, each capability component registers with the central router built into the client. The central router receives and records the identifier, function, availability status, and dependencies of each component, generating a component registration list. After receiving the signing method decision result from the server, the client parses the target signing method combination and extracts the corresponding component identifier, then sends a component call request to the central router. The central router verifies the availability and dependencies of the components based on the component registration list. If the verification is successful, the corresponding component is dynamically instantiated. Then, according to the preset execution order of the decision result, the instantiated components are orchestrated, defining the execution order, data transmission rules, and exception handling logic, generating an intelligent signing task pipeline, driving each component to complete the corresponding signing operation according to the pipeline.

8. A device for configuring personalized delivery confirmation methods, characterized in that, include: The package attribute library building module is used to build a multi-dimensional attribute library for packages; The package feature vector construction module is used to extract the package attributes corresponding to the target package based on the constructed multi-dimensional package attribute library, and construct the feature vector of the target package based on the package attributes of the target package and the logical relationship between the attributes. The signature rule matching module is used to intelligently match and decide on signature rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and the digital signature rule library.

9. An electronic device comprising a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the electronic device to perform the steps of the method for personalized configuration of receipt as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the steps of the method for personalized configuration of the receipt method as described in any one of claims 1-7.

Images